CN1688914A - Methods and systems for improved boundary contrast - Google Patents

Abstract

The present invention relates to a method and a system which use a diffractive micro-mirror array to determine the characteristic boundary of an irradiation-sensitive medium on a workpiece and then are extended to form a pattern and a structure on a semiconductor underlayer. The workpiece consists of a lithographic mask, an integrated circuit and other electronic and optical devices.

Description

Improve the method and system of boundary contrast

Technical field

The present invention relates to use the diffraction-type micro mirror array to determine the characteristic boundary of the irradiation sensitive media on workpiece, extend to the method and system that on Semiconductor substrate, forms pattern and structure.Workpiece comprises lithographic mask, integrated circuit and other electronics and optical device.

Background technology

The form that is used for forming in integrated circuit or device production two kinds of main radiation energy of pattern is photon beam and electron beam.Generally, utilize the system Billy of photon beam of a plurality of scannings with the more frequent use of the system of a plurality of electron beams.Photon or laser pattern generator system are general than very fast, but do not have electron beam system accurate.Laser beam at a plurality of broads of laser scanning system has than the different characteristic of single electron beam in the electron beam system that drives at vector.Can use some ornaments (embellishment) in writing carrying out mask with laser scanning system, partly to compensate the bigger beamwidth of photon beam.

For the application of writing direct, may more preferably use the photon exposure irradiation.Because electron beam may have adverse effect to the layer character of integrated circuit.In the substrate place and electron charge trap layer of integrated circuit, the electronics of the resistant layer (resist layer) by wanting composition may damage or change the characteristic of the layer below this resistant layer.The characteristic of these changes may have undesirable influence to device performance.Based on the writing station of photon an advantage being arranged is that generally it is faster than electron beam device.

These inventors develop the novel pattern generator that uses the photon exposure irradiation continuously.Novel pattern generator uses a micro mirror array (in one embodiment, being spatial light modulator (" SLM ")) and pulse irradiation source, replaces using one or more scanning laser beams, prints so-called mark (stamp) on surface of the work.Above-mentioned graphics engine (Graphic Engine) is with the inventor a kind of in several application that some aspects of this novel pattern generator overlap to be described.These applications of examining also illustrate the micro mirror array stamped mark that can use other kinds with pulse irradiation.

The micro mirror array of exploitation relies on diffraction, rather than deflection forms contrast on the irradiation sensitive media.By using diffraction, the small movements of micro mirror makes irradiation disperse.This disperses corresponding in an objective plane, the destructive interference from the irradiation component that a micro mirror is transmitted out.Hole and other optical elements with this dispersion change into the corresponding plane of delineation of irradiation sensitive media on workpiece in the gray scale intensity variations.

A micro mirror on objective plane can be formed on the Gaussian distribution of the light intensity in the plane of delineation.By various approximate, the roughly corresponding zone of 3 * 3 or 5 * 5 grids of the light distribution influence of a micro mirror and micro mirror.On the contrary, the light intensity of the exposure irradiation on the point in the plane of delineation depends on the orientation of 9 or 25 micro mirrors on objective plane.

Irradiation sensitive media (for example resistant layer) has certain thickness and opaque on the plane of delineation.The top of this medium is different to the response condition of exposure irradiation with the bottom.This is decided by the characteristic of this medium and will exposing and the borderline contrast between the exposed areas not.Generally, the bad meeting of contrast forms top wide and narrow base or sidewall off plumb ditch, and this is corresponding to the profile (iso-exposure profile) that exposes that waits by this dielectric thickness.Because dielectric thickness changes, particularly after the corrosion of etching process, the sidewall out of plumb boundary position of may trading off.Compare with the cd variations that allows, the variation of dielectric thickness is big.Sin10 ° or Sin15 ° is 0.1736 or 0.2588, and this expression off plumb sidewall makes the degree of position on border to the dielectric thickness sensitivity.

Might improve boundary position and/or contrast.Because have the transport function relation between the position on border and micro mirror tilt, so the working range of micro mirror can influence the position on border.Scaling method also can improve boundary position and/or contrast.

Summary of the invention

The present invention relates to use the diffraction-type micro mirror array to determine the characteristic boundary of the irradiation sensitive media on workpiece, and then extend to the method and system that on Semiconductor substrate, forms pattern and structure.Workpiece comprises lithographic mask, integrated circuit and other electronics and optical device.Concrete aspect of the present invention illustrates in instructions and the accompanying drawing at claims.

Description of drawings

Fig. 1 represents a kind of general layout of micro mirror pattern generator;

Fig. 2 represents the scope of mirror tilt;

Fig. 3 A represents the amplitude and the light intensity curve of plural number; Fig. 3 B represents can be by the complex amplitude of various mirror structures generations; Fig. 3 C-3E represents that slm pixel value and resistant layer image space coagulate as (aerialimage) corresponding with various complex amplitude scopes;

Fig. 4 represents to use negative black inclination (negative black tilt) influence to bad mirror tilt;

Fig. 5 represents the formation of side wave lobe (side lobe);

Fig. 6 represents to influence the different mirror shapes that the side wave lobe forms, array structure and surface profile;

Fig. 7 represents individually and as the part of micro mirror array, the variation of the position of the phase interference structure on mirror surface;

Fig. 8 represents transport function;

Fig. 9 is the block scheme of an optical system;

Figure 10 represents the operable various patterns of calibration technique;

Figure 11 represents the influence to local marginal position of bad edge and adjacent pixels;

Figure 12 represents the position of the line of the variation that timing signal is useful;

Figure 13 and 14 is represented the various aspects of demarcating;

Figure 15 is the process flow diagram of a calibration algorithm;

Figure 16 represents repeatedly to write (multipass writing); With

Demarcation before Figure 17 represents repeatedly to write.

Specific embodiments

Carry out following detailed description referring now to accompanying drawing.Preferred embodiment just illustrates the present invention, is not to limit its scope.Claims restriction in the scope of the present invention.Those skilled in the art can do the change of various equivalences from the following description.

Fig. 1 represents the general layout of a micro mirror array pictcure generator.In one embodiment, this generator uses a spatial light modulator (" SLM ").The various aspects of this SLM pattern generator illustrate in the above-mentioned patented claim of examining.The workpiece that exposes is placed on the worktable 112.The position of worktable is controlled by accurate localization device (for example paired interferometer 113).This workpiece can be for having the mask of a resist or other exposure sensitive material layers, perhaps for writing direct, and can be for having the integrated circuit of resist or other exposure sensitive material layers.The continuous motion on first direction of this worktable.With another direction of first direction approximate vertical on, this worktable or slowly motion, perhaps step-by-step movement motion makes the mark striped (stampstripes) that can expose on workpiece.In this embodiment, receive a flash of light instruction 108 on pulse excitation thing lasing light emitter 107, this instruction can produce laser pulse.This laser pulse can be in deep ultraviolet (DUV) or extreme ultraviolet (EUV) spectral range.Utilize beam adjuster or homogenizer, this laser pulse is converted to irradiates light 106.Beam splitter 105 guides to SLM104 with at least a portion of this irradiates light.This pulse is of short duration, (for example having only 20ns length), and therefore in the flash of light process, any motion of worktable is all freezed.The data stream 101 that the SLM104 response is handled by image raster device (pattern rasterizer) 102.In a structure, SLM has 2048 * 512 catoptrons, and each is 16 * 16 μ m, and has the projected image of 80 * 80nm.It comprises a CMOS analog memory, and this storer has on each memory node, forms a micromechanical mirror of half micron.Electrostatic force between memory node and catoptron drives this catoptron.This installs at diffraction pattern, rather than works under the direct reflection pattern, and only need make edge deflection 1/4 wavelength (being 62nm when 248nm) of this catoptron, goes to complete off-state from complete on-state.In order to form a meticulous address grid, this mirror drive to connecting, is disconnected and 63 intermediate values.Pattern by millions of images of SLM chip synthetic (stitch) together.Flash of light and synthetic carry out with the speed of 1000 mark/seconds.In order to eliminate synthetic and other errors, utilize compensation grid (offset grid) and electric field, write this pattern four times.In addition, along the edge, this electric field mixes.The single demarcation of catoptron.To be placed in the light path the light activated CCD camera of excimer, be placed on the position with image equivalence under final lens.The SLM catoptron drives by the sequence of known voltage, and this camera is measured response.Each catoptron is all determined a calibration function, be used for, proofread and correct gradation data in real time at ablation process.In data channel, the pattern grating of vector format is turned to gray level image, its grey level is corresponding in 4 times write, the radiant quantity level on the single pixel (dose level).Then, utilize image processing method, handle this image.Last step is for being the driving voltage of SLM with this image transitions.Utilize the FPGA (Field Programmable Gate Array) part to make the Flame Image Process function in real time.By each step described in the relevant patented claim, the pattern data of rasterisation is converted to the value 103 that is used to drive SLM104.

In this structure, SLM is the diffraction pattern micro-mirror device.Technically various micro-mirror devices have been announced.In another structure, irradiates light can be conducted through a miniature light gate device (for example in the LCD array) or micromechanics light valve.

As shown in Figure 2, micro mirror be can drive and " disconnection " or 1/4 wavelength delustring inclination surpassed.Fig. 2 A represents the catoptron vertical with exposure irradiation 212.In this embodiment, use beam splitter.Make the input and output of micro mirror array cross identical passage.In another embodiment, input irradiates light and mirror surface make that output can be in the channel set different with input at an angle.The complex amplitude of the output that any one point from catoptron 212 sends can be expressed as a point on the unit circle of representing its real part and imaginary part (phase place) amplitude components.In Fig. 2 A, the real part of the reference position of catoptron is+1, and imaginary part is 0.Because vertical with the input irradiates light, therefore for all points in catoptron 212, this reference position 210 is all identical.

Fig. 2 B represents to be approximately a pitch angle (inclination among the figure is illustrative, draws in proportion) of 1/3 of 1/4 wave tilt.Catoptron 222 can compare reflector position on the edge and the reflector position in the centre from the inclination of reference line 224.1/4 wave tilt is producing 180 ° phase differential between the light of catoptron edge and middle reflection.The complex amplitude of a side of catoptron 225 is expressed as an arc 221 on unit circle.221 expressions of this arc are from the center to outward flange, at the complex amplitude of the lip-deep point of catoptron 222.The complex amplitude on full reflection mirror surface is represented from second arc of arc 221 reflections with crossing X-axis.For the sake of clarity, in this array of figure, only represented a side of catoptron.

Fig. 9 represent with projection optics system described below such as same low-pass filter.In image, the physical construction of catoptron is low pass filtering device and removes.What stay is average complex amplitude on mirror surface, and this amplitude can be converted to light distribution.For a surface, the average complex amplitude on this surface can be with a surperficial integral representation:

The complex amplitude that is drawn depends on the complex amplitude of beginning.For simplifying the analysis, desirable sometimes plural reflectivity γ is 1.Refractive index h is on the edge of catoptron.Wavelength X is an irradiation mode.

Simplify to be applicable to various situations, for a common micro mirror, this surface integration ds can be reduced to a pile integration dx from the right hand edge of this catoptron to left hand edge.In some cases, the xsect of catoptron is that the function of x changes.For the rectangular mirror that does not tilt, when getting reflectivity and be 1, be reduced to:

φ in the formula ₀Angle on the-mirror surface is a constant for the catoptron with flat surface.For rectangular mirror, be reduced to

When getting reflectivity and be 1, further be reduced to

In Fig. 2 B, the real part of plural integration is represented with 220.It is in the scope of arc 221.Recall, for a centroclinal catoptron, arc 221 also exists on the catoptron of x-axle inclines face.In this case, counter-clockwise direction and clockwise direction arc have complementary imaginary part.In Fig. 2 C, represented 1/4 wave tilt.The edge of catoptron 232 is 1/4 wavelength from the distance that reference line 234 tilts.Vertical light of advancing with catoptron, on the edge that raises than the distance of lacking 1/2 wavelength of advancing at the catoptron center, the reason of Here it is generation destructive interference and diffraction.Arc 231 expression is from the contribution to the complex amplitude of catoptron of a side of catoptron.The complex amplitude 230 that is drawn is drawn in initial point, and it is consistent with destructive interference.In Fig. 2 D, catoptron 242 too tilts corresponding to so-called " negative black (negative black) " from reference line.This negative black pitch angle and is determined the contrast character on characteristic boundary by analysis, and the mirror design relation, and transport function is demarcated and the manufacturing of the semiconductor devices of following explanation all is beneficial to.Arc 241 has been described phase differential about 257 ° between the input and output ripple.This phase differential is between 1/4 wave tilt and 1/2 wave tilt.For 16 microns wide catoptrons 16, this is corresponding to tilting about 87.5 or 88nm at the edge, and by the light source irradiation of a 248nm.This is new slant range.

Fig. 3 A represents complex amplitude and light intensity curve.These curve representations are exported from the forwarding that a micro mirror sends in the middle of an array tilts identical micro mirror.Unique composition of the destructive interference of these curve representatives is the interference in the output component of a micro mirror.These curves are successively demarcated explanation in (ripple labor cut Gray demarcation) application above-mentioned ripple labor cut Gray.In this application, Fig. 3 A of similar instructions of Fig. 6.Generally, this application requires 1/4 wave tilt of certain catoptron of intersecting at the x axle corresponding to the dynamic range of Fig. 3 A curve that to set a lower border be amplitude 310 and light intensity 312.

In their work in advance, all catoptrons in a micro mirror array are demarcated in this inventor and colleague's suggestion thereof, make them produce identical maximum reflection output and identical minimal reflection output.When public minimum output, the control signal that ripple labor cut Gray demarcates drives a specific mirror tilt to a point shorter than 1/4 wave tilt.At this moment, this specific catoptron produces the obtainable low output level of all micro mirrors.Like this, can produce the dynamic range of a unanimity, and minimum output intensity can be in fact near zero.

Fig. 3 A～3B represents to have proposed different transport functions and calibration strategy in this drawing by the output of the new slant range generation of catoptron.In Fig. 2 A～2D, also represented new slant range.In Fig. 3 A, the minimum real part of complex amplitude curve 310 is below the x axle.This minimum real part is corresponding to the angle of the real part and the arc 241 of the complex amplitude among Fig. 2 D 240.A corresponding local maximum light intensity of minimum amplitude with 306 negative values is arranged in curve 308.Curve 306 available functions Sin (x)/x are approximate.Because light intensity curve be amplitude curve square, therefore, the light intensity of this part maximum is consistent with negative minimum amplitude.

Fig. 3 B～3E represents how to utilize expose contrast in the resistant layer of improving positive and negative of negative black.Fig. 3 B is the unit circle 1600 that mates with Fig. 2 A～D.Point A is corresponding to 210 of Fig. 2 A, and some B, C and D are corresponding to 220,230 and 240 among Fig. 2 B～2D.Add some E.Utilization has the tilt axis of departing from, with some mirror structure (for example below Fig. 7 C) near the relative more effective reflecting surface of tilt axis, then can design practically have be created in this above axis with the catoptron of the slant range of the complex amplitude scope of following identical positive and negative.That is: for this mirror structure, the shape of the curve 306 among Fig. 3 A on horizontal axis 302 with following identical.To this mirror structure, some A and E represent minimal tilt and maximum inclination; Perhaps Zui Da positive amplitude and minimum negative amplitude.Point A and E have identical or equal real part and identical or equal average complex amplitude under certain angle of inclination.Because according to the convention of any one specific catoptron, the value of some A gets+and 1,0, therefore to two different catoptrons, some A reflection output is not the same.

Fig. 3 C～3E comprises the pixel value with reference to the coding of the diffraction element (for example catoptron) of the point that marks on unit circle.In Fig. 3 C, the pattern of pixel value is DDAAAADD.In this coding, A is bright catoptron, and D is that complex amplitude is zero dark catoptron.In this pattern, the catoptron that switches on and off is adjacent, does not add negative black.In Fig. 3 D, pattern is BBAAAABB." B " catoptron produces negative amplitude (for example, the rectangular mirror of inclination 87nm).With positive amplitude, negative amplitude is little relatively.In Fig. 3 E, pattern is EEAAAAEE.This is a special pattern, and it is selected by having, and the catoptron that produces the slant range of positive " A " amplitude of equating and negative " E " amplitude produces.Reach in the application of the working range of this point and " high precision printing " that the catoptron characteristic proposes at the same time and be described.

Fig. 3 C representation space coagulates the influence of the DDAAAADD pattern of pixels of picture and resistant layer image.Coagulate in the picture (shown in Fig. 3 D～3E) in the space shown in Fig. 3 C, zero light intensity exposure is matrix lines or axis, activates the dotted line 301 of exposure for changing of resistant layer.The real part of corresponding complex amplitude is with (dotted line of Fig. 3 D～3E) is represented for negative sometimes.Corresponding to the complex amplitude of representing with solid line square light intensity profile be positive.In Fig. 3 C, complex amplitude is identical with light intensity curve.The relative position of three points 326 is represented the sharpness of contrast, and vertical a pile point is represented high-contrast.How these some expressions utilize the neighbor of sensor or video camera to see this image.The many more variations corresponding to the reading of being got by the neighbor of a sensor array of vertical point are big more.Point among Fig. 3 C heap vertical not as in other figure; this relatively tilts corresponding to the sidewall in the resistant layer image 322 on substrate 324; in order to represent conveniently; resistant layer image after expression is developed; but though whether mask or certain device are arranged, identical sub-image formed in any irradiation sensitive layer on substrate.

Fig. 3 D represent with more approach vertical point pile 336 corresponding, the higher contrast ratio on the sidewall of resistant layer 332.The dashed curve that picture is coagulated in the space comprises negative complex amplitude.As a result, in the zone of the resistant layer that will remove development, the light intensity curve that solid line is represented have little on the occasion of.Yet at the outer ledge place of this curve, below 301, therefore, resistant layer image 322 has desirable shape and ratio more subvertical sidewall in Fig. 3 C to little positive light intensity in the threshold value that triggers resistant layer.

Fig. 3 E is different with above-mentioned figure, because it has a negative resistant layer or irradiation sensitive layer.That is: resistant layer image 342,343 keeps its minimum exposure in threshold value below 301, rather than maximum exposure is more than threshold value.The irradiation light intensity of threshold value more than 301 triggers negative resistant layer, but by corrosion, rather than it is anticorrosive to harden, and can remove.The complex amplitude curve that dotted line is represented extends the same far with above below axis, and therefore, the light intensity curve of being represented by solid line has identical or equal height in each of three nodes.Corresponding with the more subvertical sidewall of resistant layer image 342,343, some heap 348 is than more approaching vertical at above-mentioned Fig. 4.This sidewall is approaching more vertical, and the influence of then shining the variation in thickness of sensitive layer and other variations in the developing process is more little.

Fig. 4 represents to use negative black to tilt to the influence of bad mirror tilt.This figure sets, and wishes that the inclination of catoptron is identical, but does not reach for all dark catoptrons.That is: in 402 row, the inclination of desirable 1/4 wave reflection mirror is 62nm (for the wavelength of 248nm).Yet the actual mirror tilt that obtains can be in 61～64nm scope.As a result, dark/and bright (zero tilts) catoptron between border 412 and 414 fluctuations, be not straight line.The emulation that utilizes complicated method and optical analysis software (for example SOLID-C) to carry out shows that the negative black of dark catoptron tilts to weaken the influence of bad mirror tilt.Be expert in 404, the scope of dark mirror tilt is 86～90nm.Can find out that from Fig. 3 A curve 304 is near its negative minimal value, and is more smooth with axle 302 intersections than it.Can observe the curve that the dotted line among Fig. 3 D～3E is represented equally.The positive amplitude of the negative amplitude of dark catoptron and adjacent bright catoptron interacts.When negative amplitude curve than flat the time, it is not too crucial to interact.Border 422 and 424 to 412 and 414 more straight is because they are more insensitive to bad mirror tilt.In Fig. 4, bright catoptron is near dark catoptron.In this structure, negative black tilts strong especially along the border to the influence of contrast.Obviously, attempt on marginal sharpness, to obtain 15～35% improve (for example, common United States Patent (USP) 6373619B1 number, the 6th～7 goes (on April 16th, 2002)), but nearest development is at practical problems, and the discovery contrast is more a lot of than improving of the previous imagination.One of practical problems of research is the influence (for example, the grey catoptron of the inclination of the 30nm between the bright and dark catoptron that 0～62nm tilts) at the grey catoptron between the bright and dark catoptron.The analysis showed that even when the grey catoptron occurring between bright and dark catoptron, the black of utilization tilts, can make the homogeneity of crucial size improve 2 times.The combination of the problem of the grey catoptron of amplitude of bearing very greatly and the resistant layer of bearing and centre, not explanation before in inventor's application.

Fig. 5 represents to use the side wave lobe of negative black generation.The side wave lobe is the phenomenon of discussing in the phase shift mask scope that weakens sometimes.Suppose, target 502 is exposed.Little reflection mirror array or single catoptron form the Gaussian distribution of partly representing with Figure 50 4.Curve 312 expressions of Fig. 3 A at the local maximum place, in the inclination beyond zero light intensity tilts, have remaining light intensity (for example, 4.8% of largest light intensity).According to the size of remaining light intensity, the overlapping place of remaining light intensity sending from several adjacent targets can form side wave lobe 506.For example,, go up the place of printing serif, can form the side wave lobe at some serives (serifs) proofreading and correct in order to carry out optical approximate.The size of remaining light intensity depends on the shape of the catoptron in miniature luminous lens array.

Fig. 6 represents to influence different mirror shapes, array structure and the surface profile that the side wave lobe forms.Utilize emulation catoptron 602 to show and form about 4.77% remaining light intensity.Catoptron as 604 can produce 6～8% bigger remaining light intensity.The remaining light intensity that catoptron 606 produces less than 4.77%.Use, can be reduced to remaining light intensity near 0% no matter the marking (footprint) is square or rhombus along the catoptron 608 that the axis between the relative turning pivots.Yet catoptron 608 produces than any all more difficult printing array in other confirmations.In order to reflect the irradiates light that is incident upon on the array 618 effectively, micro mirror 608 is arranged along two biasing axis 622 and 624.Utilize rhombus catoptron and two axis must get well in the method described in the graphics engine application not as array work with single group axis.

The graphical representation of the bottom of Fig. 6 can obtain the method for array shape of the rule of the little remaining light intensity of catoptron 608 and catoptron 602,604 and 606.Catoptron 32 is a rectangle, but can be another kind of regular shape.Shown in xsect, the surface of this catoptron has highly different zone 634.The part 644 of the rising shown in the xsect is approximately 1/4 ripple on the first type surface of catoptron 642.The diffraction of this damaging property of part interference and the scope by mirror tilt.The rising surface area of a unit is offset with the area of the lower surface of same size, and identical with the distance of tilt axis.Fig. 7 represents the 634 corresponding regional various potential sites that raise with Fig. 6.

Fig. 7 represents single and as the variation of the position of the phase interference structure on the mirror surface of a part of micro mirror array.In Fig. 7 A, interference structure 704 is placed on the corner of catoptron, and is diamond shaped shape 702 in the centre of catoptron.Interference structure 704 act as counteracting pattern as 708 in, two times of their surface areas in reflection output make interior zone 706 can effectively produce reflection output.Interference structure 704 produces the interior zone 706 of the reflector space of the catoptron 608 among similar Fig. 6.Use the catoptron of interference structure 704 to be rectangle, make and can form an array with catoptron, the center of the reflecting surface of this catoptron is aimed in each row and column, does not need the skew row and column on second group of axle of array shown in Figure 6 618.When forming array, roughly all elements in this array all can be by the mirror over interference structure.

Make as not the covering or open of this array of the ditch between the catoptron, need the spuious irradiates light handled to reduce than the zonule.This spuious irradiates light can be along optical channel, on undesirable direction, be reflected, absorb by this array, or damaging under the catoptron to involve the mirror drive element guiding (channe1) (in order to further specify damage to the mirror drive element, can be referring to by people such as inventor Grehinski, No. 10/338981, the common U.S. Patent application that is entitled as " using the high-energy of miniature electronic mechanical devices; the optical system of the anti-irradiation of low energy densities " that on January 2nd, 2003 proposed is introduced for reference here).Wish to use to have and show greatly that regular grids covers the interference structure of reflection mirror array and along the few catoptron of the incidence reflection of optical channel.Also wish to use to have and show regular grids greatly and cover reflection mirror array and the element below mirror surface catoptron the few interference structure of the exposure of spuious irradiates light.In Fig. 7 A, the marking of the reflecting surface of catoptron have one roughly the tilt axis at center with at the identical reflector space of the two opposite sides of tilt axis.In this embodiment, the reflector space on the two opposite sides of this tilt axis crosses this tilt axis, symmetrically reflection.At another embodiment (for example catoptron 608 of Fig. 6), but reflecting surface primary event (crossing tilt axis) then along tilt axis, or around the central point of this catoptron.Interference structure on the tilt axis two opposite sides departs from the reflection output intensity of this tilt axis, near than on the tilt axis or it little.In this embodiment, effectively reflecting surface is a triangle, and its base is along tilt axis, and its summit is being left on the edge of tilt axis.The position of interference structure is near apart from tilt axis apart from the opposite edge ratio of catoptron.Preferably, the odd-multiple of this interference structure 704 a plurality of 1/4 wavelength of level of departing from principal reflection structure 702 up or down.The odd-multiple of a plurality of 1/4 wavelength produces 180 ° of phase differential between light that reflects from this interference structure 704 and reflecting surface 702.The technician knows that the characteristic of above-mentioned Fig. 7 A can make up with many modification, to produce useful mirror structure.

Fig. 7 B represents another structure of phase shift interference structure.Near tilt axis interference structure 714 is than leaving the big of tilt axis.In this modification, this reflecting surface 712 is produced departs near the tilt axis reflection output intensity 716 (proportional with effective reflective surface area), near than on the tilt axis or it big.The twice that effective diffraction region 718 is approximately interference structure 714 is big.Effective reflector space moved near tilt axis can make minimum complex amplitude more for negative.Like this, can increase the amount of " negative black ", make contrast and edge acuity higher, more potential side wave lobe is arranged simultaneously.When this class of a curve when Fig. 3 E, the contingency of side wave lobe causes using negative resistant layer.

Fig. 7 A-7B lumps together expression, by using different interference structures on the catoptron with the identical basic marking and layout, the amount of the negative black that produces in the time of can revising mirror tilt.Do not need to change again the shape or the layout of the catoptron marking, can satisfy many requirements.

The periphery that Fig. 7 C is illustrated in reflecting surface 722 uses phase interference structure 724 on every side, to form less catoptron effectively.Effectively diffraction region 728 makes catoptron profile 726 less, and it can throw less light distribution, thereby increases spatial resolution, particularly repeatedly write fashionable especially like this.Utilize peripheral interference structure, can make the effective radius of distribution reduce half.Reducing of this usable reflection zone can be on adjacent catoptron, or on the row that replaces of catoptron, or on the catoptron checkerboard pattern repeatedly.When considering adjacent catoptron and interference structure thereof, two sides (Fig. 7 D) or two turnings (Fig. 7 E) of each catoptron can have peripheral interference structure, make reflection mirror array have complementary interference structure.This periphery interference structure can move apart the short distance at catoptron edge later on, setting up the interference effect along border between two bright catoptrons, rather than the structure that is based upon the homophase between the adjacent interference structure is disturbed (in-phase constructive interference).

One or more transport functions can be converted to mirror deflection with desirable characteristic displacement.The transport function of needs is converted to mirror tilt with characteristic displacement.Though the relation between the grey of taking a picture cannot be converted to characteristic displacement linearly, can use the shorthand of gray-scale value as characteristic displacement.In the micro mirror array hardware effort, the function of displacement and inclination and driving voltage function or other can be able to be realized the transfer function combination of desirable characteristic displacement.Fig. 8 represents a transport function.Y-axle 802 is corresponding to the edge feature displacement.In other words, whether the exposure that set (in positive medium) this medium enough in the exposure sensitive media? X-axle 84 is corresponding to mirror deflection.Curve 806 is relative to each other them.Just as was expected, and when mirror deflection reached maximum inclination, curve was more smooth and not too responsive to bad inclination.Should be appreciated that at amplitude curve 310 and this transport function above the light intensity curve 312 and extend into negative amplitude range.Different with the light intensity curve 312 of Fig. 3, this curve is reduced to the little edge dislocation at maximum inclination place monotonously from the big edge dislocation of minimal tilt.On this working range, in the micro mirror of other same tilt, in fact the output intensity of a micro mirror increases (for example, when tilting when 62nm increases to 87.6nm).Yet this function is dull, because near bright micro mirror, and the light intensity that is produced in the plane of delineation of negative black micro mirror, interference of destroyed property and the diffraction that is produced by the interaction between this negative black and the bright catoptron reduce.The edge feature displacement is to utilize negative deflection to improve a practical methods of the contrast at characteristic boundary place to the transport function of mirror deflection (or its sub, for example, deflection driven voltage).

As comprise that ripple labor cut Gray demarcates described in other the common applications of application and the patent, consider the performance of single catoptron, need to demarcate.Ongoingly may comprise new scaling method, some of them have in the catoptron of negative black maximum inclination particularly useful in demarcation.A kind of optical system that expression can be used for demarcating among Fig. 9.The technician understands that when stating scaling method in the use, the orientation of element can change, and can add or change other optical element.In Fig. 9, irradiation irradiates light 901 is incident upon on the micro mirror array 902.Generally, this irradiates light is a pulsed, can send from an impulse source, or produce by interrupting a continuous light source.On objective plane, drive this micro mirror array, to produce pattern.Guide to the surface (for example beam splitter 906) of a partial reflection from the light of this reflection mirror array reflection, and transfer to first lens arra spare or a suitable focusing mirror.A hole 908 is set on Fourier plane.This hole can be than hole 916 little (diaphragm that compares further dwindles) or bigger than hole 916.The irradiates light of exposure guides to another partial reflection surface 912 by second lens arra spare 910, a part of irradiates light is guided to workpiece 920 and a part of irradiates light guides to calibration sensor 926.The irradiates light of guiding workpiece 920 is by 914, the second holes 916 of the 3rd lens arra spare and the 4th lens arra spare 918.This light generally focuses on on the workpiece 920 one the irradiation sensitive layer.The irradiates light of guiding calibration sensor 926 is by the 5th and the 6th lens arra spare 922,924.As a kind of selection, between lens 922 and 924, can place the another one hole.Scale down on the workpiece 920 will produce than image little on calibration sensor 926.This calibration sensor can be a Charge Coupled Device (CCD) (CCD) video camera, a MOS video camera, or a charging injection device (CID, Charged Injection Device).The CCD-video camera is by Kodak ^The video camera that KAF 1600 produces has about 1000*1600 pixel and for the sensitivity of the wavelength (for example 248nm or 197nm) of use.Generally, this sensitivity comprises and utilizes fluorescent dye that irradiates light is converted to visible light, but also can use directly responsive to short wavelength (for example 248nm) camera chip.Preferably, the radiant quantity of electromagnetic radiation is approximately 0.8 times of this sensor maximum magnitude.When be incident upon on the sensor radiant quantity too hour, in some cases, signal to noise ratio (S/N ratio) is unacceptablely low.When the radiant quantity on being incident upon sensor is too big, then sensor supersaturation, outcome measurement out of true.

Before can demarcating SLM, slm pixel must be mapped on sensor array or the video camera, so that set up geometric relationship between sensor and the pixel.Utilize catoptron to generate and be generally bright or opposite image on dark background by what more coarse some grid constituted.Calculate the position of each point, and the grid of match one distortion.Utilize the figure of distortion to seek on reflection mirror array the pixel that the particular element with sensor array adapts.The further details of this process and variation are demarcated in the application ripple labor cut Gray and are provided.

Figure 10 represents the various patterns that can use in calibration technique (comprising new method).Figure 10 A represents by demarcate the pattern that so-called blanket formula (blanket) the gray scales fixed output quota described in the application is given birth to ripple labor cut Gray.In this method,, export to produce identical reflection nominally all pixels in a field are all adjusted.For example 1010 such pixels change with other pixels, and it produces different light intensity on plane of delineation part.Regulate the output of sending repeatedly, mate with adjacent pixels from the pixel that produces different light intensity.This scaling method is along curve, and the more curve in Fig. 3 B 310, rather than the transport function in curve 312 or Fig. 8 806 is followed the tracks of the pixel output intensity.

Figure 10 B represents a kind of checkerboard type scaling method.It is not to set a pixel, field to produce identical reflection output intensity, but the half-pix in the checkerboard pattern 1022 is adjusted to bright (or dark) value.Second half of pixel 1024 is changed to negative black (or bright) by a scope from a grey is arranged.This scaling method produces as the transport function among Fig. 8 806, because the interaction between the neighbor under different tilt mode influences the light intensity in the plane of delineation.

Figure 10 C represents a kind of sparse chessboard scaling method.The pixel that does not work of this pattern separately.According to a module, the pixel 1032 that does not work by three, four or five bright pixels 1034 separately.According to another module, separating between the pixel that does not work is relevant with the plane of delineation light intensity profile of a catoptron.By emulation rather than by this light intensity profile that measure to produce corresponding in the plane of delineation by the light distribution that does not have the irradiation of the interfere with or compromise of other pixels to produce accordingly with specific bright pixel.Adapt with the minimal tilt of the constraint of dynamic range used in pixel work, this bright pixel can be bright fully (for example at the minimal tilt place), or selectively bright, corresponding to the limited minimal tilt of the dynamic range of using in pixel operation.This module is the root mean square of light distribution.For this module, adopt the RMS of standard definition:

According to this standard definition, r is in a position in the plane of delineation at the centre distance r place of light distribution, and E (r) measures or the exposing radiation amount in the plane of delineation at r place for light intensity.Preferably, utilize the quadratic integral of following simplification, behind the background exposure energy that removes light source rather than bright pixel, calculate this value:

In this formula, provide (x, y) coordinate of some r.It is this center of distribution at least that the bound of integration covers, and (x0 is y0) with the scope that comprises roughly all exposing radiation amounts that bright pixel is sent.Another kind method is, this integration can utilize polar coordinates, carries out in the scope that comprises substantially all exposing radiation amounts of sending from this bright pixel.Not bright pixel be separated into the K* light intensity RMS that center to center is measured, if make that the center of its light distribution separates the distance of 2* light intensity RMS on the plane of delineation, then a pair of not bright pixel separates 2* light intensity RMS.The separation of bright pixel is not more than or equal to 2* light intensity RMS, 3* light intensity RMS, or 4* light intensity RMS.2* light intensity RMS, 3* light intensity RMS or 4* light intensity RMS module are compared with bright catoptron computation measure standard, (the micro-mirror structure of Fig. 7 C～7E) for example that can be used to have peripheral interference structure better.Preferably, do not work separation between the catoptron more than or equal to 4* light intensity RMS.Be separated into 2* light intensity RMS or 3* light intensity RMS is not so good.These separate to simplify greatly and isolate a specific catoptron that do not work and be incident upon the light distribution on the plane of delineation or the work of effect.In sparse checkerboard pattern, the response of single catoptron can be demarcated by grey to negative black from bright.Sparse checkerboard pattern can be around the skew of tiny mirror array, till all single catoptrons are all demarcated.Need to use than the more sparse checkerboard pattern of checkerboard pattern in order to demarcate, because in sparse chessboard, just having in 16 pixels has one for not working in one or 25 pixels, rather than in normal chessboard, has one to be the pixel that does not work in two.The pattern that the number that uses is many can be required the not interactional iterations between the bright pixel that isolation separates 4* light intensity RMS each other reduce compensate.

Figure 10 D～10E represents another kind of scaling method, and wherein, dark pixel band is across being inswept on the bright pixel field on border with the pixel that does not work.Figure 10 A～10E can be used for illustrating improved scaling method.Micro mirror and more generally speaking is any SLM element of modulated complex amplitude or light intensity, can utilize the light intensity that produces in by sensor or camera acquisition image to demarcate.In Figure 10 A, generally by all elements being carried out identical modulation (making alive), can first approximation ground with all mirror drive to grey uniformly.Measuring the actual ash value that produces is (for example pixel 1010) how to change on array on the image that video camera produces with it, and the information of the modulation that improves particular element can be provided.As demarcating described in the application ripple labor cut Gray, this process can be improved this and be similar to till producing good uniform grey camera review repeatedly.Recording voltage figure is as calibration maps.After demarcating several ash values, this figure is compressed into the catoptron Parameter Map.

It is bright that improved photometry calibration process is used some catoptrons, the pattern that other do not work for dark formula, and measure resulting grey level, rather than with the extremely uniform grey of mirror drive.A kind of improved method uses bright checkerboard pattern (Figure 10 B) with the element that does not work to demarcate dark value for based on the uniform gray image that is produced.Another improved method is used the background of a color (for example bright or dark), and with the element drives of isolating to another color (for example not working or non-dark (Figure 10 C)), and measure and demarcate the ash value that the element by this isolation produces.These improved scaling methods are at the end of dynamic range, and particularly the dark end in dynamic range can provide result preferably.

In lithographic system, the photometry character of catoptron is coupled with printing character, but these character itself are inessential.Light intensity is important, because it influences the position at the edge in the image.Shown in Figure 10 D～10E, improved demarcation can comprise and drive the micro mirror array with line pattern, and along bright and dark between the boundary demarcation pixel.

For the idealized system with equal catoptron, if the input data are the straight edge 1101 among Fig. 1, then photometry is demarcated and is produced straight edge.Utilize the catoptron 1100 on real border (real-world), existence can not be demarcated the unfavorable condition of eliminating fully.This photometry scaling method stays remaining ripple 1102 on the edge.Because the rectilinearity at edge and correctness are important print parameters, also be valuable even therefore in edge control, improve a little.

The line pattern data of packing in the modulator.Produce image from this modulator, and call wire pattern image 1102.The position and the rectilinearity at line edge are quantized, and according to the line position 1105 on the ad-hoc location 1104 at this edge, calculate the correction to adjacent pixels 1106,1107,1108.Shown in the line position among Figure 12, be based on the specific pixel and near line position and linear to the correction of specific pixel.Added correction can be the weighted mean corresponding to several corrections of the line position of the variation among Figure 12.Preferably, at least one on the transport function 806 of correction chart 8 (preferably 2 or a plurality of) point.Itself can use this method repeatedly, perhaps demarcates the another kind of method of using next time.

Process is as follows: a dense line intermittent pattern with a parallel axes is added on the SLM.In a preferred embodiment, spacing is 7 pixels, and therefore, bright and dark zone is 3.5 and 3.5 pixel wide.For the sake of clarity, Figure 11～Figure 14 has only represented an edge of line intermittent pattern.SLM is illuminated under the condition identical with the condition that is used to write actual pattern, and the irradiates light of projection is by translucent light beam sampling thief or beam splitter 912 taps, so that form second image 1109 on sensor array 926.Formed second image with reach workpiece on the exposure sensitive media on image identical, but ratio is adjusted to the resolution of sensor array 1109, and the introducing noise 1110 relevant with sensor.The image that captures that integrates to reduce camera noise and optics spot is stored in the computing machine that is connected, and analyzes.

Before catching image, should utilize photometry and geometry method to demarcate this sensor array, the geometry demarcation can comprise exports to SLM and computed image distortion pattern with a sparse dot pattern.Photometry is demarcated and can be utilized even photograph to carry out.Simultaneously, can measure other errors (for example dark current in sensor array) and tabulation.

Can utilize the nominal data of storage, the geometric configuration and the sensitivity of the image of being caught by sensor array are proofreaied and correct.The image border that analysis captures.Marginal position 1105 in center pixel 1103 is subjected to the influence of all catoptrons in radius (for example 1104).(shown in this radius be similar to, for various light intensity RMS factors coupling, can determine various radiuses).Bright neighbor 1107 influences pixel 1104 with its minimal tilt.Dark pixel 1105 is exerted one's influence with its maximum inclination.Gray pixels 1108 is exerted one's influence with its medium pitch.In principle, can be that predetermined level is found edge 1101 by making threshold value.In fact, available mathematical method finds and has the line edge preferred with the match of the spatial frequency of optical resolution 1301 compatibilities (can suppress produced by camera pixel 1110 and other stochastic sources 1300 scrambling).Site error 1304 along this this edge of edge calculations.As shown in figure 13, calculate the correction of neighbor, and accumulate the correction of getting up to be used to calculate adjacent pixels.According to weighting function, will on adjacent pixels, expand in the correction of a position.

Just the pixel on error vector 1304 is maximum to its corrected value contribution, and immediate neighbor contribution is less, and the contribution of second neighbor is just very a little bit smaller.Weighting function can pre-determine, and can also determine by rule of thumb or improvement in calibration process.In little neighborhood, not only can provide relative sensitivity, and can provide with for example digital value of the reality of nm/DAC value representation when changing the voltage on a catoptron and writing down moving of edge.Corrected value can be used DAC value representation and certainty ratio, makes when process finishes, and the long-pending corrected value in boundary is the desirable correction with the DAC value representation.

A pixel can be given more than 1 corrected value.For example, as shown in figure 14, can calculate relative amplitude for+1 ,+0.4 and-0.1 corrected value.In Figure 14, catoptron has three corrected values being realized by totalizer (accumulator).When marginal position as shown in the bottom right time, the marginal position of pixel 5 is subjected to the white value of pixel 1 and 4, the influence of the dark value of pixel 2 and 5 intermediate value and pixel 3 and 6.The correction of calculating is weighted and adds up in accumulator register.As shown in figure 12, this line intermittent pattern is offset 11/2 pixel, and writes down and handle new image.Another kind of scheme is at record with before handling a new images, can make this line intermittent pattern be offset another suitable distance.When one of this line intermittent pattern skew in a small amount the time, can directly calculate the more point on transport function 806.When crossing the spacing of 7 pixels, during 0.5 pixel of this line intermittent pattern skew, before recording new image of record, a complete set of all half-integer location of pixels that 14 images will be this patterns.This pattern can be turned to different orientation (for example, from vertical rotation to level), and record and handle the new collection (for example, 14 images) of image.Also can write down 4 groups of orientations (for example, level, vertical and two diagonal line).Can demarcate the geometric configuration that will print, making has most important line orientation in a line intermittent pattern, and writes down and handle many array images.After obtaining and handling all images, or become the catoptron correction chart of correction, and can proofread and correct the catoptron table with white, black and intermediate value.

In a preferred embodiment, according to the difference of counting in proofreading and correct pattern, can demarcate catoptron in advance with light measuring method, and be offset by the voltage ratio to each catoptron, certainty ratio or extension can obtain new correction pattern.For 3 or more point, extension can utilize smooth function to carry out.For three points, quadratic polynomial can match.For more than three points, a cubic spline function passes through calculation level.Can regulate this transport function, make it and near maximum inclination, adapt with a more smooth transport function.As mentioned above, the preferred embodiment is three points.(for example n) can make the displacement between each image less more than three points, is generally (1/ (n-1)).Principle does not change, and the scope of displacement is widened.

Figure 15 is the process flow diagram that aforesaid edge is demarcated, and expression multi-stage iteration (potentialiteration).In this process flow diagram, at first in 1151, determine weighting function.Yet as mentioned above, weighting function can be determined by rule of thumb on other positions of this flowchart process, perhaps improve during the course.Trimming process begins from replacement totalizer 1512, and these totalizers can reset to zero, a predetermined value, or pre-determine or the value of the single pixel that iteration is determined.By from vector format to the rasterisation form, with a line gratingization, can in 1513, produce a line pattern to be added on the micro mirror array.Another kind method is, the data of the simple pattern that is used to demarcate by storage can not need rasterisation, demarcate logical circuit and can produce the line of demarcating usefulness.In 1514, image projection is listed at a sensor array.In 1515, detect the edge in the projects images.In 1516, for example, filter out camera noise by to time integral or average a plurality of image.In 1517, determine along the site error of a position at edge.In 1518, this error is weighted function distributes, and is accumulated in the correction totalizer of neighbor.One skilled in the art will appreciate that step 1517 and 1518 can change and resequence, make and, before entering corresponding totalizer, to determine a plurality of site errors and to calculate a correction according to weighting a plurality of site errors.Carry out multi-stage iteration.Though be shown in iteration on edge orientation (1502) marginal position (1503) and the edge pixel (1504) by the sequence list of determining, the order of iteration can change.Can differentiate complex pattern fast in real time and pattern is delivered to the processor system of micro mirror array, also can be apace according to edge orientation, any order of position and pixel produces pattern.In 1501, outside iteration loop represents that repeatedly this process is till it converges to satisfied result.In addition, though weighting function 1511 in iteration loop 1501 outsides, can be in processing procedure, determine by rule of thumb or improve weighting function and in fact carry out the iteration of arbitrary number of level.

Precision according to the storage correction factor can be added in error diffusion on the non-integral correction factor.If it is a DAC unit that a catoptron distributes the correction of 0.53DAC unit and storage precision, then can obtain the correction of a complete unit, and error ε=0.53-1.00 can spread between immediate neighbor or scatter.Like this, can reduce error source and digital noise.

Marginal position is demarcated the photographic system that can be used for any SLM and produce line.No matter SLM is between white and null black (for example A and the D among Fig. 3 B), perhaps drive in grey and negative deceiving between (for example A and the B among Fig. 3 B), implementation all is identical, bigger dynamic range makes more wishes to use a plurality of three calibration points, or pre-determines the curve characteristic of a flex point.

As the dynamic amplitudes scope of the SLM of strong phase shift mask be from-1 pass through 0 to+1 (E Fig. 3 B for example, D, A).Therefore, some most probable is approaching+1, another point approaching-1, the three o'clock near 0.Shown in Fig. 3 E, the interference between the light intensity that the complex amplitude by negative, positive produces will form two edges 348 in negative resistant layer, rather than will form an edge 326,336 in positive resistant layer.Need Flame Image Process to change a little, go to survey two edges, rather than an edge.For example, the pixel of a plurality of 1 medium pitch can be placed between the pixel of minimal tilt and maximum inclination, their the two because relation of light intensity, if amplitude square can produce brightness from the complex amplitude of positive and negative, then all be " bright ".

Above-mentioned and described in the claims method can be used to demarcate various pattern generators based on array, the example of the array that can demarcate is for based on direct reflection (TI for example, the described mirror design of Daewoo) the grid light valve (for example, Silicon Light Machines company), the LCD modulator element, (for example absorb array, Hank Smith MIT), with based on electrooptics (for example Xerox), photoelasticity (photeelastic), acoustooptics (acustooptic), magnetooptics, (for example LEDs, VCSELs) optical array of character or SLM interferometry or emission.Light can utilize the single lens of refraction or diffraction, (for example, Hank Smith, MIT) imaging of a plurality of lens or lens arra.Other arrays are the near field optic array, and machinery (for example IBM) and electric (for example Carbon Naontubes) writes array.In addition, can be to based on modulator array, for example (for example, SPIE 2002 for orifice plate, Canon), photoemitter (for example, Mapper) or a plurality of particle post (particle column, for example, ETEL Broody), uses these scaling methods to ion or electron beam.For optical writer, the preferred video camera of using for image pickup that uses.For particle beam, can use the device (for example particle detector array) of measuring particle.No matter use the exposure energy and the modulator of what form, always can expose resistant layer and inspection exposing patterns, as using sensor or video camera to collect the alternative method of real time data.

For demarcation, that need cannot not be staggered according to this image or staggered, adopting diverse ways, the micro mirror SLM of diffraction forms contiguous not images interlaced with partially coherent light.The state of the vicinity of image and influence reflection modulator array.Other forms of array modulator (the lens arra optical system of for example nano-tube array, or MIT) produces images interlaced.Utilize non-conterminous modulator element, and form adjacent pixels in different time.The staggered calibration software that makes becomes complicated, but does not change above-mentioned principle, and calibration software need be known staggered character and weighted correction is added on the suitable correction totalizer.

Above-mentioned scaling method can extend to repeatedly and write.Purpose is to demarcate this array, makes the error that do not have that writes.In the system of any reality, error residual value (error residual) is arranged all.Shown in Figure 16 A～16C, when print pattern repeatedly, this residual value can reduce by equalization.Twice printing can make stochastic error reduce 2 square root.Yet, as shown in figure 17, owing to can measure and know error.And compensate between each time, therefore the error residual value is reduced better than random fashion.And for example shown in Figure 17ly can change process, making can be to determining weighting function for adjacent pixels in other times, and can accumulate the error of these errors.Repeatedly proofread and correct more simply but be not effectively process for will be in write-once the residual value tabulation of each pixel.Utilize the skew (offset) of known field and pixel then, calculate error repeatedly.Handle the table of corrections of whole tiny mirror arrays again, and to pixel correction residual value repeatedly.

As if a kind of saying is arranged, and repeatedly correction can be added to big computation burden on the scaling method.In fact, the calculating strength the best part of this method is the marginal position in the abstract image.Though the thin note (book-keeping) that repeatedly writes is complicated, the difference when applying the correction work based on one or many is little.

From above-mentioned explanation, the technician knows, becomes part with array from each side of the present invention, can construct a large amount of various system and methods.Embodiment is for using the inclination micro mirror array of two dimension, to the method for definite characteristic boundary of at least one the irradiation sensitive media exposure on the workpiece.In addition, this method can be used for the micro mirror array of the inclination of one dimension.This method comprises makes first group of micro mirror tilt in an example on described border, produce the output of high reflection, and the opposite side of second group of micro mirror on described border tilted, by producing the inclination of minimum reflection output intensity, to the inclination that produces the contrast of obviously improving along the border.This high reflection output can be subjected to and will be exported by the maximum that micro mirror obtains, by with the available value corresponding dynamic of various micro mirrors scope in this array, or the restriction of selecteed dynamic range, this selecteed dynamic range is used to increase by the positive complex amplitude of first group of micro mirror generation and the ratio of the negative complex amplitude that is produced by second group of micro mirror.The complex amplitude of the reflection output that draws from second group of micro mirror can have the real part of the complex amplitude of the negative value of being roughly.According to the dynamic range of the selection of height reflection output and the structure of micro mirror, the ratio of the absolute value of the negative complex amplitude of a specific micro mirror and positive complex amplitude can be approximately 0.218,0.5 even be 1.0 the order of magnitude.According to the difference of applicable cases, the ratio of hope can more than or equal to 0.5, or be approximately 1.0 more than or equal to 0.2.For example, the application of mask manufacturing can be used less ratio, and the application of writing direct can be used bigger ratio.Determine along the border the obviously contrast of improvement by emulation or by sub-image or the developed image estimated on resistant layer or another irradiation sensitive media.An aspect of this method can be to select the inclination of first and second groups of micro mirrors, makes the destructive interference of respectively organizing between the micro mirror improve contrast greatly.The 3rd group of micro mirror can be placed between first and second groups of micro mirrors, and the 3rd group of micro mirror has medium inclination.The inclination of the 3rd group of micro mirror can produce the gray scale that adapts with monotonic quantity.The scope of this monotonic quantity can reflect the output that exports to by the inclination generation of second group of catoptron from this height.This monotonic quantity can make at the edge dislocation of irradiation on the sensitive media relevant with inclination on the micro mirror edge.Can reason out from the curve shown in the figure, the inclination of second group of micro mirror of flex point distance of this monotonic quantity, nearer than the inclination of first group of micro mirror of distance.Another aspect of this method comprises how micro mirror tilts.They can be around roughly at the axis tilt at center, or from lopsidedness.In the described deformable micromirror device of Texas Instruments, tilt to comprise the supporting member distortion that makes micro mirror, perhaps can comprise making micro mirror distortion itself.

The various aspects of said method can further tilt to make up with making this micro mirror repeatedly.The irradiates light of the available partial coherence of this inclination shines this reflection mirror array and guides this reflection output to form at least one pattern and processing workpiece on the irradiation sensitive media, forms and the corresponding one or more semiconductor structures of this pattern on workpiece.When this workpiece was a graticule (reticle), above-mentioned method can make up: this micro mirror that tilts repeatedly, shine this reflection mirror array and will reflect output with the irradiates light of partial coherence and guide on this irradiation sensitive media; Developing pattern on this graticule; With the one or more and corresponding semiconductor structure of this pattern of formation on Semiconductor substrate.

When another embodiment shines the sensitive media exposure for making on workpiece when the inclination micro mirror array that utilizes one dimension or two dimension, determine the method for a characteristic boundary.This embodiment can comprise makes the lopsidedness of first group of micro mirror on this border, produce the output of a high reflection, tilt with the opposite side that makes second group of micro mirror on this border, by in the reflection output of second group of micro mirror, producing maximum destructive interference, reach the inclination that between the reflection output of first and second groups of micro mirrors, produces tangible destructive interference.When same embodiment exposes for make irradiation sensitive media on the workpiece when the inclination micro mirror array that utilizes one dimension or two dimension, determine the method for a characteristic boundary.This same embodiment comprises makes the lopsidedness of first group of micro mirror on described border, produce the output of high reflection, with the opposite side of second group of micro mirror on described border tilted, after the inclination that produces minimum reflection output intensity, reach the inclination that between the reflection output of first and second groups of micro mirrors, produces tangible destructive interference.Tangible destructive interference in these embodiments is also more than misalignment or the wrong destructive interference that causes accidentally of demarcating owing to catoptron.Significantly destructive interference can or be estimated the pattern that produces and determine by emulation in the irradiation sensitive media.The various aspects of first embodiment and further combination can with any combination among these two embodiment.

The modification of the embodiment of several means and this method and said method adapts.The embodiment of a device determines the one dimension of characteristic boundary or the controller of two-dimentional inclination micro mirror array for when exposing this irradiation sensitive media on workpiece.This controller comprises logical circuit and the source that can be operatively connected with this reflection mirror array, this controller can drive first group of micro mirror on one side of border to first inclination, and the high reflection output of generation, and can drive second group of micro mirror on the opposite side of border to second inclination, by producing the point of minimal reflection output intensity, to producing the point that obviously improves, in this embodiment along the contrast on this border, the variation of the inclination of second group of micro mirror can be parallel with the variation in the said method.This embodiment is connected with controller and is subjected to the inclination micro mirror of this controller control can comprise the reflecting surface marking with one or more following features: one roughly at the tilt axis at center, similar reflective surface area on the opposite side of this tilt axis, with depart from this tilt axis, than near the reflective surface area that on this tilt axis or it, reduces (or increase) greatly.Above-mentioned controller embodiment also can comprise the pattern generator parts.These pattern generator parts can comprise the radiation source of a projection illumination on this array.Optical system will transfer to the irradiation sensitive media on the workpiece from the irradiates light of this reflection mirror array reflection.In addition, form the worktable of a holding workpieces, the may command workpiece motion s is to limit characteristic boundary.Another scheme is, adopts or do not adopt the pattern generator parts can use this device to remove to drive the 3rd group of micro mirror between first and second groups of micro mirrors to medium inclination.This medium pitch can be determined by the transport function of dullness.The scope of this monotonic quantity can comprise inclination that produces high reflection output and the inclination that produces the contrast of obviously improving.The flex point of this dullness transport function produces the inclination of the contrast of obviously improving from this, than nearer from the inclination that produces high reflection output.

The product of making embodiment comprises the medium that can be read by machine.This medium has when utilizing micro mirror array to make irradiation sensitive media when exposure on the workpiece, and the indication of characteristic boundary is determined in control, also comprises the indication of the modification that realizes above-mentioned any method and method.

The present invention also comprises several embodiment of inclination micro mirror.The embodiment of an inclination micro mirror that can use in micro mirror array or in fact use comprises the marking of a reflecting surface.The marking of this reflecting surface has roughly the tilt axis at the center, in the similar reflective surface area of the two opposite sides of this sloping shaft with leave this tilt axis than on this tilt axis or near the reflective surface area of little it (or big).This similar reflective surface area laterally is symmetrical (for example mirror image) this tilt axis, perhaps the lateral symmetry of the point that they can be on this tilt axis (for example, secondary crosses the image of vertical axis reflection).

Another embodiment is the reflecting surface of the inclination micro mirror that can use in micro mirror array or in fact use.This reflecting surface comprises a tilt axis that has roughly at the center, in the similar reflective surface area of the two opposite sides of this sloping shaft with leave this tilt axis than on this tilt axis or near a reflecting surface marking of the reflective surface area of little it (or big).In this embodiment, the slant range of this reflecting surface comprises first pitch angle and second pitch angle that produces high reflection output, this second pitch angle is beyond the inclination that produces minimum reflection output intensity, to the inclination that improves contrast greatly that produces between the micro mirror with first and second pitch angle.Tilt angle ranges in this embodiment can be used above-mentioned method representation.

Another embodiment also comprises can be at the reflecting surface of micro mirror array use or the inclination micro mirror that in fact uses.This reflecting surface has roughly the tilt axis at the center, with a marking of reflective surface area similar on the opposite side of this tilt axis.This reflecting surface marking also has the phase interference structure on the two opposite sides of tilt axis, and this interference structure makes this reflecting surface compare little (or big) on the sloping shaft or near it at the reflection output intensity that leaves this tilt axis.The aspect of this embodiment can for, this reflecting surface marking is included in the opposed edges on the centroclinal axis two opposite sides, and more close this opposite edge of phase interference structure and this tilt axis.Another aspect of this embodiment can be included in opposed edges on the end opposite of centroclinal axis for, this reflecting surface marking, and this interference structure is nearer than the mid point between this opposite edge from this opposite edge.Another aspect of this embodiment can for, this interference structure is along the outward flange of this reflecting surface marking.Can reduce the size of the projection light distribution of micro mirror effectively along this outer peripheral interference structure.With any one embodiment of the reflecting surface of this inclination micro mirror, the height of this phase interference structure and this reflecting surface itself differ odd-multiple 1/4 wavelength.

Another embodiment is for demarcating the method for two-dimentional micro mirror array.This method can be worked with first and second parallel axes ground, but also can be only along an axis job.This method comprises parallel with first axle, is created at least one the first contrast lines between the bright and dark micro mirror.At least be on some line positions, the micro mirror array with grey value that this first contrast line crosses between dark and bright micro mirror applies.Say that in logic this line can cross on the entire emission lens array by inswept.Yet the pattern generating logical circuit can make this first line at random be positioned at each position on this reflection mirror array.Record reflects output intensity accordingly with this micro mirror on each frontline position.Preferably, make the process that produces the first contrast line, repeatedly for the second contrast line of second parallel axes parallel with the discord first axle.Say that in logic first axle can pass the center of micro mirror, and second axis also can pass the center of micro mirror, if arrangement of micro mirrors becomes cartesian array, then this axis can be vertical.Yet, also can adopt other orientations of axis.The process that produces the contrast line can be to the 3rd and the 4th axis on diagonal line repeatedly.This method also comprises carries out and the corresponding calculating of proofreading and correct of the reflection output intensity of record single micro mirror.The aspect of this embodiment is, first and second contrast lines between bright and dark micro mirror comprise the dark micro mirror of at least three polyphones, the bright micro mirror of the micro mirror that these are dark or contiguous these at least three polyphones is perhaps separated by the bright micro mirror of at least one grey micro mirror with these three polyphones at least.When this method of use, apply the contrast line and can comprise this contrast line of placement, at least three record and each contrast lines from the reflection output intensity of a pixel output are adapted.

Another embodiment is for demarcating the method for a two-dimentional micro mirror array.This micro mirror array is by the partial coherence light source irradiation of an electromagnetic radiation with characteristic wavelength.This method is included in the checkerboard pattern that produces micro mirror in this reflection mirror array, alternately is set at bright the square of this chessboard and the value of not working.The square of this chessboard can be a single pixel or a rickle pixel (for example, 2 * 2 or 3 * 3 pixels).This method also is included in the tilt axis and the scope between 1/4 and 1/2 wavelength difference between the edge of micro mirror, drives the micro mirror that does not work, with the light intensity of exporting on the document image plane.It also comprises one or more motivation values of determining this micro mirror that does not work, be used to set can with the response curve of the motivation value coupling of the output intensity of two records.The output intensity of this record is the representative of edge dislocation.The aspect of this embodiment is according to the motivation value of determining, sets the dark and dynamic range of the output intensity of reflection mirror array.

Another embodiment is the scaling method of two-dimentional micro mirror array that utilizes the electromagnetic radiation source irradiation of partial coherence.This method is included in and produces a micro mirror pattern in this reflection mirror array.Most of micro mirror is set at first output intensity.Other micro mirror is the scope by the output intensity value just in time.Other micro mirror is roughly spaced-apart.This method comprises that these other micro mirrors of driving to the scope of output intensity value with under various drive signals, write down the output intensity of other micro mirrors.Then, determine the drive signal of single micro mirror, to produce desirable output intensity level.This drive signal can realize with calibration curve or transport function.The aspect of this embodiment can be at least three micro mirrors of first output intensity value for separating in other micro mirrors.In addition, separating between the micro mirror center of measuring on the plane of delineation can be for corresponding with general bright micro mirror, the root mean square factor of light distribution.As mentioned above, separating between the micro mirror center can measure on the plane of delineation.This factor can for light distribution all the side 2,3 or 4 times.

Another embodiment is the method for the two-dimentional micro mirror array of demarcation exposure irradiation light source.This method is included in the pattern that produces light source in this reflection mirror array.A wherein big light source is set at the scope that first output intensity value and other light sources are set at the output intensity value.Other light sources fully is separated from each other, and can individually be differentiated by sensor.This method comprises the driving other light sources by output intensity value scope, and writes down the output intensity of other light sources under various drive signals.Determine the drive signal of light source, to produce desirable output intensity level.Another aspect of the present invention is to determine drive signal for single source.

The present invention also comprises the logical circuit and the source of realizing above-mentioned any method.It can extend to the pattern generator that comprises this logical circuit and source.As a kind of product of manufacturing, it also comprises the storer with the DLC (digital logic circuit) that realizes above-mentioned any method.It also extends to the pattern generator of the DLC (digital logic circuit) that manufactures a product of packing into.

Though with reference to preferred embodiment and the example that describes in detail above the present invention has been described, these examples are illustrative purposes, not restriction.The technician can be used for various modifications and makes up, and these improvement and combination are all in the scope of material of the present invention and following claims.

Claims (42)

1. work as the inclination micro mirror array that utilizes a two dimension for one kind, when making at least one the irradiation sensitive media exposure on the workpiece, determine the method for a characteristic boundary, this method comprises:

Make the lopsidedness of one first group of micro mirror, to produce high reflection output on described border;

The opposite side of one second group of micro mirror on described border tilted, after the inclination that produces minimum reflection output intensity, the inclination of the contrast of obvious improvement to generation along this border.

2. the method for claim 1 is characterized by, and described generation improves greatly between the reflection output of the described first and second groups of micro mirrors of being tilted in of contrast and produces destructive interference.

3. the method for claim 1 is characterized by, and it also is included between described first and second groups of micro mirrors, makes one the 3rd group of micro mirror tilt to medium pitch.

4. method as claimed in claim 3, it is characterized by, described the 3rd group of mirror tilt produced and the corresponding gray scale of monotonic quantity, and this monotonic quantity comprises from producing the obviously scope of the inclination of the contrast of improvement of described generation that tilts to of high reflection output.

5. the method for claim 1 is characterized by, described micro mirror round one roughly at the axis tilt at center.

6. the method for claim 1 is characterized by, and described micro mirror is from lopsidedness.

7. the method for claim 1 is characterized by, and described inclination comprises the supporting member distortion that makes described micro mirror.

8. the method for claim 1 is characterized by, and described inclination comprises makes the micro mirror distortion.

9. the method for claim 1 is characterized by, and the inclination of the contrast that described generation obviously improves is corresponding to the negative value that is roughly of the real part of the complex amplitude of the reflection output of sending from described second group of micro mirror.

10. the method for claim 1 is characterized by, and this workpiece is a Semiconductor substrate; Described method also comprises:

This micro mirror is tilted, utilize the irradiates light of partial coherence to shine this reflection mirror array and guiding reflection output at least one pattern of formation on this irradiation sensitive media; With

Handle this workpiece, on this workpiece, form and the corresponding one or more semiconductor structures of this pattern.

11. the method for claim 1 is characterized by, this workpiece is at least one graticule, and described method also comprises:

This micro mirror is tilted, utilize the irradiates light of partial coherence to shine this reflection mirror array and will reflect to export and guide on this irradiation sensitive media;

Be developed in the pattern on this graticule; With

On Semiconductor substrate, form and the corresponding one or more semiconductor structures of this pattern.

12. work as the inclination micro mirror array that utilizes a two dimension for one kind, when making at least one the irradiation sensitive media exposure on the workpiece, determine the method for a characteristic boundary, this method comprises:

Make the lopsidedness of one first group of micro mirror, to produce high reflection output on described border;

The opposite side of one second group of micro mirror on described border tilted, through maximum destructive interference in the reflection output that is created in described second group of micro mirror, to the inclination that between the reflection output of described first and second groups of micro mirrors, produces tangible destructive interference.

13. work as the inclination micro mirror array that utilizes a two dimension for one kind, when making at least one the irradiation sensitive media exposure on the workpiece, determine the method for a characteristic boundary, this method comprises:

Make the lopsidedness of one first group of micro mirror, to produce high reflection output on described border;

The opposite side of one second group of micro mirror on described border tilted, through producing the inclination of minimum reflection output intensity, to the inclination of the tangible destructive interference between the reflection output that is created in described first and second array mirrors.

14. a controller that is used for determining the two dimension inclination micro mirror array of characteristic boundary when the irradiation sensitive media of exposure on workpiece, this controller comprises:

Logical circuit and the source that is connected with this reflection mirror array can be used for:

Drive one first group of micro mirror in the lopsidedness on described border and the output of the high reflection of generation; With

Drive the opposite side of one second group of micro mirror and tilt, through producing the point of minimal reflection output intensity, the point of the contrast of obvious improvement to generation along the border on described border.

15. inclination micro mirror as claimed in claim 14 is characterized by, it is connected with this controller and controlled by this controller, and it comprises a reflecting surface marking, and this marking has:

Roughly at the tilt axis at center;

Similar reflective surface area on this tilt axis opposite side; With

Leave this tilt axis, than near reflector space much smaller on this tilt axis or it.

16. inclination micro mirror as claimed in claim 14 is characterized by, it is connected with this controller and controlled by this controller, and it comprises a reflecting surface marking, and this marking has:

Roughly at the tilt axis at center;

Similar reflective surface area on this tilt axis opposite side; With

Leave this tilt axis, than near reflector space much bigger on this tilt axis or it.

17. a pattern generator, it comprises the described controller of claim 14, also comprises:

Irradiates light is incident upon a irradiation source on the reflection mirror array;

To transfer to the optical system on the irradiation sensitive media on the workpiece from the irradiates light of this reflection mirror array reflection; With

When determining characteristic boundary, holding workpieces, the worktable of controlled motion.

18. controller as claimed in claim 14, it is characterized by, this logical circuit and source also are used between first and second groups of micro mirrors, comprising the inclination that produces high reflection output and producing in the scope of inclination of the contrast of obviously improving, drive the 3rd group of micro mirror to a corresponding medium pitch of monotonic quantity.

19. a pattern generator, it comprises the described controller of claim 18, also comprises:

Irradiates light is incident upon a irradiation source on the reflection mirror array;

To transfer to the optical system on the irradiation sensitive media on the workpiece from the irradiates light of this reflection mirror array reflection; With

When determining characteristic boundary, holding workpieces, the worktable of controlled motion.

20. the product of a manufacturing, it comprises:

The medium that machine can be read, with when the irradiation sensitive media that utilizes micro mirror array exposure on workpiece, the indication of the definite characteristic boundary of control also comprises following indication on it:

Make the lopsidedness of one first group of micro mirror, to produce high reflection output on described border;

The opposite side of one second group of micro mirror on described border tilted, through producing the inclination of minimum reflection output intensity, the inclination of the contrast of obvious improvement to generation along this border.

21. an inclination micro mirror that is used in the micro mirror array, it comprises a reflecting surface marking, and this marking has:

Roughly at the tilt axis at center;

Similar reflective surface area on this tilt axis opposite side; With

Leave this tilt axis, than near reflector space much smaller on this tilt axis or it.

22. an inclination micro mirror that is used in the micro mirror array, it comprises a reflecting surface marking, and this marking has:

Roughly at the tilt axis at center;

Similar reflective surface area on this tilt axis opposite side; With

Leave this tilt axis, than near reflector space much bigger on this tilt axis or it.

23. a reflecting surface that is used in the micro mirror of the inclination in the micro mirror array, it comprises: a reflecting surface marking, and this marking has:

Roughly at the tilt axis at center;

Similar reflective surface area on this tilt axis opposite side; With

Leave this tilt axis, than near reflector space much smaller on this tilt axis or it;

Wherein the slant range of this reflecting surface comprises:

Produce one first pitch angle of high reflection output; With

One second pitch angle, this angle surpasses the inclination that produces the minimal reflection output intensity, between the micro mirror with described first and second pitch angle, produces the inclination of the contrast of obviously improving.

24. a reflecting surface that is used in the micro mirror of the inclination in the micro mirror array, it comprises: a reflecting surface marking, and this marking has:

Roughly at the tilt axis at center;

Similar reflective surface area on this tilt axis opposite side; With

Phase interference structure on the opposite side of this tilt axis, the position of this structure make this reflecting surface leave this tilt axis, than on this tilt axis or near produce it reflection output intensity little.

25. reflecting surface as claimed in claim 24 is characterized by, this reflecting surface marking also is included in the opposed edges on the opposite side of centroclinal axis, and this phase interference structure from this opposed edges than nearer from this tilt axis.

26. reflecting surface as claimed in claim 24 is characterized by, this reflecting surface marking is arranged to make the center of reflecting surface to embark on journey to aim at and row are aimed at.

27. reflecting surface as claimed in claim 24 is characterized by, the height of this phase interference structure and this reflecting surface itself differ the odd-multiple of a plurality of 1/4 wavelength.

28. a method of demarcating the micro mirror array of two dimension, it comprises:

Parallel with first axle, between bright and dark micro mirror, produce at least one the first contrast lines; Between this dark bright micro mirror, at least some first-line positions, utilize the micro mirror of ash value, apply the first contrast line across this reflection mirror array, and record and this micro mirror reflect output intensity accordingly;

With the discord second parallel parallel axes of first axle, between bright and dark micro mirror, produce at least one the second contrast lines; On the position that is some second lines at least, between dark and bright micro mirror, utilize the micro mirror of ash value, apply the second contrast line across this reflection mirror array; And the record and this micro mirror reflect output intensity accordingly; With

Calculate correction with the corresponding single micro mirror of reflection output intensity that writes down.

29. method as claimed in claim 28, it is characterized by, the width of the first and second contrast lines between bright and dark micro mirror comprises the dark micro mirror of at least three polyphones, and these catoptrons separate with the bright micro mirror vicinity of three polyphones or by the bright micro mirror of at least one grey micro mirror and these three polyphones at least at least.

30. method as claimed in claim 28 is characterized by, and applies the contrast line across this reflection mirror array and comprises the contrast line is set, making is that three records of the reflection output intensity that sends from a pixel are corresponding with each bar contrast line at least.

31. the scaling method by the two-dimentional micro mirror array of the electromagnetic radiation source irradiation of the partial coherence with characteristic wavelength, it comprises:

In this reflection mirror array, produce the checkerboard pattern of micro mirror, make the square of chessboard alternately be set at bright and value that do not work;

Between the tilt axis and edge of micro mirror, in the scope between 1/4 and 1/2 wavelength difference, drive the micro mirror that this does not work, and be recorded in the light intensity of the output on the plane of delineation; With

Determine one or more motivation values of the micro mirror that this does not work, be used to set the response curve of the output intensity coupling that makes motivation value and record.

32. method as claimed in claim 31 is characterized by, it also comprises, according to this motivation value of determining, sets the dark end of dynamic range of the output intensity of reflection mirror array.

33. the scaling method by the two-dimentional micro mirror array of the electromagnetic radiation source irradiation of partial coherence, it comprises:

Produce a micro mirror pattern in this reflection mirror array, most of micro mirrors are set at first output intensity value, and other micro mirrors are by output intensity value range, and other micro mirrors are roughly spaced-apart;

Drive the scope of these other micro mirrors, and under various drive signals, write down the output intensity of other micro mirrors by the output intensity value;

Determine the drive signal of single micro mirror, to produce desirable output intensity level.

34. method as claimed in claim 33 is characterized by, at least three micro mirrors that are divided into the first output intensity value in other micro mirrors.

35. method as claimed in claim 33 is characterized by, and is divided into 2* light intensity RMS at least between the center of other micro mirrors, light intensity RMS is the root mean square with the common bright corresponding light distribution of micro mirror.

36. method as claimed in claim 33 is characterized by, and is divided into 3* light intensity RMS at least between the center of other micro mirrors, light intensity RMS is the root mean square with the common bright corresponding light distribution of micro mirror.

37. method as claimed in claim 33 is characterized by, and is divided into 4* light intensity RMS at least between the center of other micro mirrors, light intensity RMS is the root mean square with the common bright corresponding light distribution of micro mirror.

38. method as claimed in claim 33 is characterized by, and drives this other micro mirrors, by producing the inclination of the minimal reflection output intensity that sends from single micro mirror, to the inclination that is created in the contrast of obviously improving between the bright and dark micro mirror.

39. method as claimed in claim 33, it is characterized by, drive this other micro mirrors, by the inclination of destructive interference maximum in the reflection output that is created in other micro mirrors, to the significantly inclination of destructive interference between the reflection output of the micro mirror with first output intensity value and other micro mirrors.

40. method as claimed in claim 33, it is characterized by, drive this other micro mirrors, after the inclination that produces the minimal reflection output intensity that sends from single micro mirror, to the significantly inclination of destructive interference between the reflection output of the micro mirror with first output intensity value and other micro mirrors.

41. a method of demarcating two-dimentional exposure irradiation source array, it comprises:

Produce a source pattern in this reflection mirror array, most of sources are set at the first output intensity value, and other sources make other sources fully spaced-apart so that by a single resolution of sensor by output intensity value range;

Drive the scope of these other sources, and under various drive signals, write down the output intensity in other sources by the output intensity value;

Determine the drive signal in source, to produce desirable output intensity level.

42. method as claimed in claim 41 is characterized by, and determines the drive signal of single source.

Images