WO2001026134A1

WO2001026134A1 - Array of multiple charged particle beamlet emitting columns

Info

Publication number: WO2001026134A1
Application number: PCT/US2000/040809
Authority: WO
Inventors: Marian Mankos; Tai-Hon P. Chang; Kim Y. Lee; C. Neil Berglund
Original assignee: Etec Systems, Inc.
Priority date: 1999-09-30
Filing date: 2000-08-31
Publication date: 2001-04-12
Also published as: KR20010089522A; AU1960601A; JP2003511855A; EP1141995A1

Abstract

A lithographic apparatus using an array of charged particle (electron) beam columns, where the array includes a plurality of charged particle beam columns that each selectively expose a target to a plurality of charged particle beams. Each charged particle beam column includes a beam source that selectively generates a plurality of charged particle beams; an anode coaxial with the charged particle beams and that accelerates the plurality of charged particle beams from the beam source; and a lens coaxial with the charged particle beams and that demagnifies the charged particle beams. The beam source is a photocathode array that selectively supplies multiple electron beams when illuminated.

Description

ARRAY OF MULTIPLE CHARGED PARTICLE BEAMLET EMITTING COLUMNS

1. Field of the Invention This invention relates to electron beam sources and, more particularly, to generation of multiple electron beams .

2. Related Art High resolution electron beam sources are used in systems such as scanning electron microscopes, defect detection instruments, VLSI testing equipment, and electron beam (e-beam) lithography. In general, e-beam systems include an electron beam source and electron optics. The electrons are accelerated from the source and focused to define an image at a target. These systems typically utilize a physically small electron source having a high brightness.

Improvements in optical lithography techniques in recent years have enabled a considerable decrease in the linewidths of circuit elements in integrated circuits. Optical methods, however, will soon reach their resolution limits. Production of smaller line- width circuit elements (i.e., those less than about .1 μm) will require new techniques such as X-ray or e-beam lithography.

In e-beam lithography, a controllable source of electrons is desired. A photocathode used to produce an array of patterned e-beams is shown in FIG. 1. U.S. Patent 5,684,360 to Baum et al , "Electron Sources Utilizing Negative Electron Affinity Photocathodes with Ultra-Small Emission Areas," herein incorporated by reference m its entirety, describes a patterned photocathode system of this type.

FIG. 1 shows a photocathode array 100 with three photocathodes 110 comprising a transparent substrate 101 and a photoemission layer 102. The photocathode is back-illuminated with light beams 103 which are focused on photoemission layer 102 at irradiation region 105.

As a result of the back-illumination onto photoemission layer 102, electron beams 104 are generated at an emission region 108 opposite each irradiation region 105. Other systems have been designed where the photoemitter is front-illuminated, i.e. the light beam is incident on the same side of the photoemitter from which the electron beam is emitted.

Often, light beams 103 or electron beams 104 are masked. In FIG. 1, light beams 103 are masked using mask 106 which allows light onto irradiation spots 108 but prevents light from being incident on other areas of photoemission layer 102. FIG. 1 also shows mask 107 which allows electrons to exit photoemission layer 102 only at certain surface spots corresponding to emission regions 105. A photocathode may also have a mask between transparent substrate 101 and photoemission layer 102 to block light beam 103 so that it is only incident at irradiation spots 105. In general, photocathode 110 may include no masking layers or may have one or more masking layers.

Each irradiation region 105 may be a single circular spot representing a pixel of a larger shape, the larger shape being formed by the conglomerate of a large number of photocathodes 110 m photocathode array 100. In that case, irradiation region 105 may be as small as is possible given the wavelength of the light beam incident on photocathode 100. Typically, a grouping of pixel irradiation regions has dimensions of 100-200 μm. Each pixel can have dimensions (i.e. diameter) as low as 0.1 μm. Alternatively, irradiation spot 105 and emission region 108 can be a larger shape. In either case, the image formed by emission region 108 will be transferred to e-beam 104 so long as the entirety of irradiation region 105 is illuminated by light beam 103.

Photoemission layer 102 is made from any material that emits electrons when irradiated with light. These materials include metallic films (gold, aluminum, etc.) and, m the case of negative affinity (NEA) photocathodes, semiconductor materials (especially III- V compounds such as gallium arsenide) . Photoemission layers m negative electron affinity photocathodes are discussed in Baum (U.S. Patent 5,684,360). When irradiated with photons having energy greater than the work function of the material, photoemission layer 102 emits electrons. Typically, photoemission layer 102 is grounded so that electrons are replenished. Photoemission layer 102 may also be shaped at emission region 108 m order to provide better irradiation control of the beam of electrons emitted from emission region 108. Further control of the e-beam is provided m an evacuated column as shown in FIG. 2. Light beams 103 usually originate at a laser but may also originate at a lamp such as a UV lamp. The laser or lamp output is typically split into several beams m order to illuminate each of focal points 105. A set of parallel light beams 103 can be created using a single laser and a beam splitter. The parallel light beams may also originate at a single UV source. Alternatively, the entire photoemission array 100 may be illuminated if the light source has sufficient intensity.

Photons m light beam 103 have an energy of at least the work function of photoemission layer 102. The intensity of light beam 103 relates to the number of electrons generated at focal point 105 and is therefore related to the number of electrons emitted from emission region 108. Photoemission layer 102 is thin enough and the energy of the photons in light beam 103 is great enough that a significant number of electrons generated at irradiation region 103 migrate and are ultimately emitted from emission layer 108.

Transparent substrate 101 is transparent to the light beam and structurally sound enough to support the photocathode device within an electron beam column which may be a conventional column or a microcolumn. Transparent substrate 101 may also be shaped at the surface where light beams 103 are incident m order to provide focusing lenses for light beams 103. Typically, transparent substrate 101 is a glass although other substrate materials such as sapphire or fused silica are also used.

If mask 106 is present either on the surface of transparent substrate 101 or deposited between transparent substrate 101 and photoemission layer 102, it is opaque to light beam 103. If mask 107 is present, it absorbs electrons thereby preventing their release from emission region 108. Mask 107 may further provide an electrical ground for photoemission layer 102 provided that mask 107 is conducting. Photocathode 100 may be incorporated within a conventional electron beam column or a microcolumn. Information relating to the workings of a microcolumn, m general, is given m the following articles and patents: "Experimental Evaluation of a 20x20 mm Footprint Microcolumn," by E. Kratschmer et al . , Journal of Vacuum Science Technology Bulletin 14(6), pp. 3792-96, Nov. /Dec. 1996; "Electron Beam Technology - SEM to Microcolumn," by T.H.P. Chang et al . , Microelectronic Engineering 32, pp. 113-130, 1996; "Electron Beam Microcolumn Technology And Applications," by T.H.P. Chang et al . , Electron-Beam Sources and Charged-Particle Optics, SPIE Vol. 2522, pp. 4-12, 1995; "Lens and Deflector Design for Microcolumns, " by M.G.R. Thomson and T.H.P. Chang,

Journal . of Vacuum Science Technology Bulletin 13(6), pp. 2445-49, Nov. /Dec. 1995; "Miniature Schottky Electron Source," by H.S. Kim et al . , Journal of Vacuum Science Technology Bulletin 13(6), pp. 2468-72, Nov. /Dec. 1995; "Electron-beam Microcolumns for

Lithography and Related Applications" by T.H.P Chang et al . , Journal of Vacuum Science Technology Bulletin 14(6), pp. 3774-80, Nov. /Dec. 1996; U.S. Pat. No. 5,122,663 to Chang et al . ; and U.S. Pat. No. 5,155,412 to Chang et al . , all of which are incorporated herein by reference in their entirety.

FIG. 2 shows a typical electron beam column 200 using photocathode array 100 as an electron source. Column 200 is enclosed within an evacuated column chamber (not shown) . Photocathode array 100 may be completely closed within the evacuated column chamber or transparent substrate 101 may form a window to the vacuum chamber through which light beams 103 gain access from outside the vacuum chamber. Electron beams 104 are emitted from emission region 108 into the evacuated column chamber and carry an image of emission region 108. Electron beam 104 may be further shaped oy other components of column 200. Electron beams 104 are accelerated between photocathode array 100 and anode 201 by a voltage supplied between anode 201 and photoemission layer 102. The voltage between photocathode array 100 and anode 201, created by power supply 208 (housed outside of the vacuum chamber) , is typically a few kilovolts to a few tens of kilovolts. The electron beam then passes through electron lens 204 that focuses the electron beam onto limiting aperture 202. Limiting aperture 202 blocks those components of the electron beams that have a larger emission solid angle than desired. Electron lens 205 refocuses the electron beam. Electronic lenses 204 and 205 focus and demagnify the image carried by the electron beam onto target 207. Deflector 203 causes the electron beam to laterally shift, allowing control over the location of the image carried by the electron beam on a target 207.

In 0.1 μm lithography systems, the size of a circular pixel incident on target 207 is on the order of 0.05 μm. Therefore, the image of emission area 108 needs to be reduced by roughly a factor of 2 to 10, depending on the size of emission region 108. Target 207 may be a semiconductor wafer or a mask blank.

Conventional variable shaped electron beam lithography columns shape the electron beam by deflecting the electron beam across one or more shaping apertures. The resulting image in the shaped electron beam is then transferred to target 207 with a large total linear column demagmfication. The requirement of large total linear demagnification (supplied by electron lenses 204 and 205) results m large column lengths, increasing electron-electron interactions that ultimately limit the electron current density of the column. The low electron current density results in a low throughput when the column is used m lithography.

Another major drawback m using known e-beam systems include the inability to modulate the electron beam without modulating the light source itself, usually a laser. Modulating a laser typically involves a large amount of control circuitry, requiring a large amount of space, and can be slow. In addition, m a patterned array of photocathodes, modulation of individual photocathodes m the array is extremely difficult. Finally, better resolution is required of lithography systems in order to meet future demands of semiconductor materials processing.

SUMMARY One embodiment of the present invention includes an array of charged particle beam columns, where each charged particle beam column selectively exposes a target to multiple charged particle beams. Each charged particle beam column includes a beam source that selectively generates a plurality of charged particle beams; an anode coaxial with the charged particle beams that accelerates the plurality of charged particle beams from the beam source; and a lens coaxial with the charged particle beams that demagnifles the charged particle beams.

In one embodiment, the beam source is a photocathode array that selectively supplies multiple electron beams when illuminated. In this embodiment, the photocathode array includes at least one photocathode, where the at least one photocathode includes a gate electrode that modulates the electron beam in response to a voltage applied to the gate electrode; and at least one pad, where the at least one pad is electrically connected to the gate electrode of the at least one photocathode by an interconnect line, and the at least one pad allows external control of the gate electrode of the at least one photocathode.

Thereby, one embodiment of the present invention includes a method of imaging a pattern onto a target, the method including the acts of: generating a plurality of charged particle beams; aggregrating the beams into groups; selectively controlling each of the groups of charged particle beams; and directing each of the groups onto the target. In this embodiment, each of the groups is contained in one charged particle beam column.

The invention and its various embodiments are further discussed along with the following figures and the accompanying text.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 shows a patterned photocathode according to the prior art . FIG. 2 shows a conventional electron beam column using the photocathode shown in FIG. 1.

FIGs. 3A and 3B show a photocathode according to the present invention.

FIG. 4 shows a portion of a photocathode array having two photocathodes according to the present invention.

FIG. 5 shows a photocathode according to the present invention having a gate electrode with multiple segments . FIG. 6A shows a photocathode according to the present invention having multiple independent segments m the gate electrode.

FIGs. 6B and 6C show sample patterned e-beams resulting from selectively turning on the segments shown m the gate electrode of FIG. 6A.

FIG. 7A epιcts the process of forming a photocathode according to the embodiment of the invention presented m FIG. 4. FIG. 8 shows a photocathode array according to the present invention.

FIG. 9 shows a micro-column utilizing a photocathode according to the present invention.

FIG. 10 shows a multiple segment gated photocathode used in an electron beam column where the beam shaping is accomplished at the photocathode.

FIG. 11 shows a conventional variable shaped beam electron beam column having multiple shaping components . FIG. 12A depicts multiple microcolumns arranged m an array, in accordance with one embodiment.

FIG. 12B schematically depicts a plan view of the array of FIG. 12A, in accordance with one embodiment.

In the figures, components having the same or similar functions are identically labeled.

DETAILED DESCRIPTION

FIGs. 3A and 3B show m a side view an embodiment of a photocathode 300 according to the present invention. (The conventional associated housing, electrical leads, etc. are not shown.) In FIG. 3A, a photoemitter 302 is deposited on a transparent substrate 301. Transparent substrate 301 is usually glass, fused silica or sapphire, although other transparent materials having structural strength sufficient for support can be used.

A light beam 303 is incident on transparent substrate 301, passes through transparent substrate 301, and is absorbed by photoemitter 302 at irradiation region 308. Photoemitter 302 emits electrons from emission area 305, located on the surface of photoemitter 302 opposite of irradiation region 308, when light beam 303 is incident upon irradiation region 308.

Emission area 305 can, m general, be of any shape and any size where gate electrode 307 determines the electric field. Some useful shapes include a circle, a square, a rectangle, an octagon and a hexagon. Irradiation region 308 should at least cover emission area 3C-..

A gate insulator 306 is deposited on photoemitter 302 such that emission area 305 is surrounded, but not covered, by gate insulator 306. Gate insulator 306 may be made from any electrically insulating material and is preferably made from Sι0₂. Gate electrode 307 is deposited on the side of gate insulator 306 away from emission region 305. Gate electrode 307 can be made from any conducting material. Photoemitter 302 can be made from any material that emits electrons when illuminated. The most efficient photoemittmg materials include semiconductors, such as cesium tellurium (Cs₂Te) , and metals, such as gold, aluminum, and carbide materials. In addition, many III-V semiconductors, such as gallium arsenide (GaAs) , are suitable photoemitter materials. Preferably, photoemitter 302 has a thickness of about Photoemitter 302 will have a work function that is determined by the actual photoemitter material. The work function is the minimum energy required to release an electron from the material . The photons in light beam 303 must have an energy at least as great as the work function in order that photoemitter 302 will emit electrons .

Light beam 303 is absorbed by photoemitter 302 at, or nearly at, the surface of photoemitter 302 corresponding to irradiation region 308. At that point, electrons will have a kinetic energy equal to the photon energy minus the work function. These electrons migrate from irradiation region 308 to emission area 305 and are emitted from the material at emission area 305 provided that the electrons have not lost too much energy to collisions within the photoemitter material. As such, the thickness of photoemitter 302 should be sufficient to absorb light beam 303 but not so thick as to reabsorb a significant number of the free electrons created.

It is also desirable that, in embodiments of this invention, the kinetic energies of the emitted electrons not be too great, preferably less than 0.5 eV but can be as great as a few eV, so that the emitted electrons can be reflected by a voltage applied to gate electrode 307. If photoemitter 302 has a work function of 3.5 to 4.5 eV, then a light beam having a photon wavelength of 257 nm or less is needed to produce photons having an appropriate photon energy. Transparent substrate 301 must be transparent to light beam 303 so that the maximum amount of light possible is incident on irradiation region 308. Transparent substrate 301 can be of any thickness but preferably is a few millimeters thick. In addition, light beam 303 may be focused to cover irradiation spot 308 in an area corresponding to emission region 305. The intensity distribution of light beam 303 is generally gaussian in shape, therefore light beam 303 will be more intense at its center than at its edges. Light beam 303 is preferably focused in such a way that its intensity is nearly uniform across irradiation region 308 so that electron beam 304 has nearly uniform intensity. In general, however, light beam 303 can be as focused as is desired.

Gate electrodes 307 are fabricated on insulators 306 and can be constructed from any conducting material. The thickness of gate insulator 306 is preferably about 1000 to 5000 A and the thickness of gate electrode 307 is also preferably about 1000 A. In one embodiment, photoemitter 302 is held at ground voltage and gate electrode 307 is biased at a voltage greater than ground, approximately +10 V, in order to accelerate the electrons that are emitted from photoemitter 302. When gate electrode 307 is biased at voltages less than ground, approximately -10V, the emitted electrons are reflected back towards photoemitter 302. Moreover, stable emission can be achieved by coupling a resistor 311 between photoemitter 302 and gate electrode 307 and using the emission-intensity for feed-back (i.e., a self-biasing system) . For example, when electron emission increases the gate voltage decreases correspondingly which in turn lowers the emission intensity. Anode electrode 310 is held at a voltage of from a few kilovolts to several tens of kilovolts and accelerates the electrons out of photocathode 300 and into an evacuated electron beam column. Alternatively, photoemitter 302 is held at a high negative voltage, gate electrodes 307 are biased at ±10 V compared to photoemitter 302, and anode electrode 310 is grounded.

In FIG. 3A, gate electrode 307 is held at +10 V. This voltage is chosen so as to be consistent with the electric field which would be set up between anode electrode 310 and photoemitter 302 if insulator 306 and gate electrode 307 were absent. With the voltage of gate electrode at 10 V, electron beam 304, which carries the image of emission region 305, is accelerated out of emission region 305. Insulators 306 and gate electrode 307 also act as a mask m order to better shape the image of emission region 305 contained m electron beam 304.

In FIG. 3B, gate electrode 307 is held at -10V. At this voltage, the electrons emitted by emission region 305 are accelerated back towards emission region 305 by the electric field created between gate electrode 307 and photoemitter 302. No electron beam

304 is created because the electrons emitted from emission region 305 are reflected back into photoemitter 302 rather than being accelerated away from photoemitter 302. Instead of an electron beam, electron cloud 309 is created where electrons are emitted out of photoemitter 302 and promptly accelerated back into photoemitter 302.

In some embodiments of the invention, the voltage at gate electrode 307 is varied m order to control the intensity of the electron beam. The higher the voltage difference between gate electrode 307 and photoemitter 302 the greater the number of electrons that leave photocathode 300. The maximum number of electrons available, those that are emitted from emission region

305 as a result of light beam 303, are extracted when the gate electrodes are set at full on (about 10V) . Although the examples shown here have the gate biasing voltage at +10V for full-on operation and -10V for full -off operation, other parameters for gate voltages are possible. The full -on bias voltage and the voltage applied to anode electrode 310 determines the thickness of insulator 306 because the electric field created by gate electrode 307 at the full-on bias voltage should be consistent with that field which would exist m the absence of gate electrode 307 and gate insulator 306. The full-off bias voltage limits the incident light beam photon energy because in full- off operation the electrons emitted from emission region 308 must be reflected back into photoemitter 302. In addition, gate electrode 307 should be the dominant feature determining the electric fields near emission region 308. The size of emission region 308 is therefore limited by the relative sizes and distances between gate electrodes 306 and anode electrode 310. In embodiments where switching times are important, the RC time constant of gate electrode 307 should be relatively small . The spacing between gate electrodes, the spacing between the gate electrode and the photoemitter, and the thickness of the electrodes determine the RC time constant and therefore the maximum rate of switching.

FIG. 4 shows an embodiment of a photocathode array 400. In FIG. 4, two emission regions 402 of photocathode array 400 are shown and both emission regions 402 are illuminated by light beam 403 which simultaneously illuminates the entire portion of photocathode array 400 shown. Parallel light beans 403 could be used instead with each beam being focused on an individual emission region 402. Photocathode array 400 comprises a transparent substrate 401, a conductor 408, gate insulator 406, gate electrodes 407, and photoemitters 402.

Conductor 408, which can be made from any conducting material but is preferably aluminum, is deposited on transparent substrate 401 and has an opening within which photoemitter 402 is fabricated. Photoemitter 402 may be any material which emits electrons when illuminated with photons, as was previously discussed. Again, photoemitter 402 has a work function and the photons m light beam 403 must have an energy at least as great as the work function m order that electrons are emitted from emission region 405. Transparent substrate 401 is preferably glass but can be any material that is transparent to light b.3am 403 such as sapphire or fused silica. Conductor 408 is opaque to light beam 403 and does not emit electrons from its front when illuminated from the back by light beam 403. Conductor 408, therefore, acts as a mask and defines an irradiation region 408.

Emission region 405 lies directly opposite irradiation region 408 on photoemitter 402 and can be of any shape and any size where gate electrodes 407 determine the electric field. A gate insulator 406 is fabricated on conductor

408 and has an opening 410 such that photoemitters 402 are not covered by insulators 406. Gate electrodes 407 are deposited on gate insulator 406. In this embodiment, gate electrodes 407 overhang opening 410 by an amount sufficient to cause the electric fields created at emission area 405 to be not substantially distorted by gate insulator 406. As m FIGs. 3A and 3B, gate electrode 407 has the ability to turn electron beam 404 on and off with a voltage applied to gate electrode 407. The on and off voltage roughly correspond to +10V and -10V, respectively. In addition, ultimate electron beam intensity may be regulated by varying the gate electrode voltage. A self-biasmg resistor 411 also may be connected between gate electrode 407 and conductor 408 m order to provide feedback for controlling the intensity of electron beam 404 by self-biasmg.

In the photocathodes shown m FIGs. 3A, 3B and 4, the intensity of the electron beams may be controlled by controlling the actual voltage between the gate electrode and the photoemitter. The lower the voltage, the less intensity that the electron beam will have because fewer of the electrons will escape the electron cloud where the electrons have a statistical distribution of velocities m the direction of electron beam propagation. In addition, the gate electrodes may be used to regulate the intensity of the resulting electron beam. In some embodiments, a resistor is placed between the gate electrode and the photoemitter so that a self-bias g feedback is created, i.e., if emission increases, the gate voltage lowers correspondingly .

In FIG. 4, gate electrode 407 is shown as being the same for each emission area 405. However, general each emission area 405 has a gate electrode 407 that is electrically isolated from the other gate electrodes. In addition, a gate electrode for a particular emission area may include several segments each of which are electrically isolated from all of the others .

In some embodiments, the gate electrode surrounding the emission region has multiple segments. Multiple segments allow the ability to turn on parts of the emission region while turning off other parts of the emission region, shaping the image carried by electron beam 404.

FIG. 5 shows a photocathode as FIG 3 but with a right gate segment 510 and a left gate segment 511 instead of single segment gate electrode 307. The result of this construction is that the electron beam can be selectively switched on. For example, FIG. 5 right gate segment 510 is held full-on at a bias voltage of 10 V and left gate segment 511 is held full- off at a bias voltage of -10 V. The resulting electric field reflects electrons which are emitted by emission region 305 near left gate segment 511 while accelerating electrons are emitted out from emission region 305 near to right gate segment 510 of photocathode 500. The resulting electron beam 504 is an image of, m this example, half the emission region 305. The resulting electron beam 504 distribution is not uniform and is most intense near right gate segment 510 and is essentially off at a point midway between the two segments 510 and 511.

FIG. 6A shows in a plan view a four segment gate electrode configuration. The gate segments are segment A 601, B 602, C 603, and D 604. Emission region 305 this example is a square. Emission region 305 can be of any shape but is preferably a square. Other useful shapes include a circle, a rectangle, an octagon and a hexagon.

FIG. 6B shows a plan view electron beam 504 that results when gate segments A 601, C 603, and D 604 are turned on (i.e., held at +10V) and gate segment B 602 is turned off (i.e., held at -10V). FIG. 6C shows a plan view electron beam 504 that results when gate segments A 601 and D 604 are turned on while gate segments B 602 and C 603 are turned off. Other shaped electron beams can be formed by selectively controlling the voltages of the segments of the gate electrodes. This ability lends great versatility to constructing photocathode arrays that are useable for a variety of different tasks. FIG. 10 shows a photocathode having a segmented gate electrode used an electron beam column for electron beam lithography.

In general , any number of gate segments can be used. The more gate segments there are, the more control a user of the photocathode has over the electron beam created from a given emission area. This ability may be of great importance in efficiently writing features onto semiconductor substrates. In addition, resistors can be coupled between individual segments of the gate electrode and the photoemitter m order to provide self-biasmg control over electron beam intensity as described above.

The photocathodes described above are conducive to miniaturization and precise integration into multiple photocathode sources . A photocathode array can be constructed on a single substrate with precise positioning of photocathodes. In particular, FIGs. 7A- 7F illustrate a process of manufacturing the photocathode illustrated in FIG. 4 using conventional semiconductor processing steps. The illustrated process shows only a single photocathode of the photocathode array. However, one skilled m the art can produce a photocathode array having precisely placed photocathodes with various emission area shapes and gate structures from this illustration. In addition, one skilled the art can modify this process order to manufacture other photocathodes according to this invention or alter this process m ways that result m the same photocathode construction.

FIG 7A shows m a cross sectional view the first step in the process where an opaque layer of conducting film is deposited on a transparent substrate 401 such as glass, fused silica, or sapphire. Preferably, transparent substrate 401 is a glass substrate. As shown m FIG. 7B, the conducting film is masked and a window having an appropriate size and shape to form an emission area 410 is etched through the conducting film. A gate insulator 406 is then deposited on top of conducting film 408 and also fills the window of emission area 410. Gate insulator 406 can be any electrical insulator but preferably is Sι0₂. A gate electrode layer 407 is then deposited on top of gate insulator 406 as shown m FIG. 7D .

Gate insulator 406 is then masked and a hole 411 is etched through gate electrode layer 407 and insulating film 406 as is shown m FIG. 7E. Hole 411 is aligned with emission area 410 and is slightly larger than emission area 410. In addition, all of insulating film 406 is removed from the window of emission area 410 by this etch.

In FIG. 7F, a selective isotropic etch has created a recessed hole 412 in insulating film 406 so that gate electrode 407 now overhangs the opening created at hole 411 and recessed hole 412. Finally, photocathode material 402 is deposited using a directional deposition technique such as thermal evaporation from a point source or ionized sputter deposition. This final deposition forms a photocathode 400 with a self-aligned gate aperture and is formed such that the photocathode is electrically connected to conducting layer 408 but maintains electrical isolation from gate electrodes 407.

In addition, m an array of photocathodes manufactured by this process, each gate electrode segment surrounding each of the photocathodes may be formed by appropriately masking the gate insulator 406 during deposition of gate electrode layer 407. Alternatively, gate electrode layer 407 may be individually etched to form individual gate segments. Also, interconnect lines that connected gate electrode segments to pads can be formed along with the gate electrode segments or may be deposited at a later process step.

As an alternative manufacturing method, the substrate could be coated with conducting layer 408, gate insulator 406 and gate electrode 407 first. Window 411 is then etched through all films down to transparent substrate 401. Using a selective isotropic etch, the opening gate electrode 407 could be enlarged slightly with respect to the corresponding window 410 m conducting layer 408. Also alternatively, multiple segments of gate electrodes are created around each of holes 411 by lsotropically etching insulating breaks m gate electrode 407. In some embodiments, the surface of substrate 401 may be shaped m order to focus the light beam onto an irradiation region corresponding to emission area 410 of photoemitter 402. Also, in some embodiments, photoemitter 402 may itself be shaped so as to better focus the resulting electron beam that is emitted from the photocathode.

FIG. 8 shows in a plan view a four by four array of patterned photocathodes. Emission areas 801 this example are squares although any shape, including circles, rectangles, hexagons and octagons, can be faoricated. Gate electrode 804 fully surrounds each emission area 801. Although only a single segment gate electrode is shown m FIG. 8, gate electrode 804 may in general be constructed of multiple electrode segments for further control of the electron emission from emission area 801. Gate electrode 804 is connected to a bonding pad 803 by an interconnect line 802. Both bonding pad 803 and interconnect 802 are preferably made from the same material as is gate electrode 804 but any conductor making electrical contact with gate electrode 804 can be used. In general, for lithography systems it is desirable that the physical separation between two adjacent emission regions be such that the array is a square. The minimum separation between emission regions is approximately four times the physical dimensions of the emission region. In FIG. 8, the dimension of the square emission region with current microfabπcation technology can be as small as 0.1 μm. Preferably, the side dimension of the emission region is 0.1 μm. Therefore, the whole four by four array shown m FIG. 8 is constructable within a square 1.6 μm on a side, which is well withm conventional microfabrication limits. FIG. 9 shows m a side view a photocathode array 910 according to this invention mounted within a microcolumn 900. Microcolumn 900 is contained within an evacuated chamber (not shown) . The substrate of photocathode array 910 may suffice as a vacuum window allowing a laser light source onto the irradiation regions of photocathode array 910 or alternatively photocathode array 910 may be fully enclosed in the vacuum chamber. Electron beams 911 are emitted from tne emission regions of photocathode array 910 and, depending on the control inputs to gate electrodes 909 of photocathode array 910, are accelerated through anode 901. Anode 901 is held at a voltage of from one kilovolt to several tens of kilovolts over that of the photoemitters m photocathode 910. Limiting aperture 902 clocks a portion of beams 911 which have a larger emission solid angle than desired. Deflector 903 allows the image of the emission regions contained m electron beams 911 to be scanned across the substrate and laterally shifted. Emzel lens, having electrodes 904, 905, and 906, focuses and demagnifles the image onto target 907. Target 907 may be either a semiconductor wafer or a mask blank for electron beam lithography.

Photocathode array 910 can include any number of individual photocathodes . Each of the individual photocathodes can include a single segment gate or a multiple segment gate. The image formed m electron beam 911 is dependent upon the emission areas of each of the individual photocathodes and the states of the gate electrodes of each of the individual photocathodes. For example, a photocathode array 910 having one photocathode with a single segment gate can only produce an image of the emission area of the photocathode. With a photocathode array 910 having multiple photocathodes, each with an individually controlled single segment gate, various images can be formed by selectively turning on the individually controlled photocathodes to form conglomerates of the images of each of the emission areas of the "on" photocathodes. A photocathode array 910 where some of the photocathodes have multisegmented gate electrodes have the most versatility because images can be formed using portions of emission areas of the individual photocathodes .

In one embodiment, limiting aperture 902 is not used so that the entire emitted solid angle of beam 911 is used. Instead, gate electrodes 909 are used to blank beam 911, i.e., prevent exposure of the target 907 from the beam 911. Accordingly the length of microcolumn 900 is reduced and thus electron-electron interactions are reduced. Such electron-electron interactions cause blurring of an image formed by beams 911 on target 907.

FIG. 10 shows an electron source 1001 having a single photocathode 1004. Photocathode 1004 has an emission area 1002 and a four segment gate structure 1003. The four segment gate structure is capable of selectively imaging emission area 1002. In the example of FIG. 10, the four segment gate structure 1003 is used to shape an electron beam image equivalent to one half of emission area 1002. The electron beam carrying the electron beam image is extracted out of photocathode 1004 by extraction electrode 1005. Demagnification lens 1006 demagnifies the electron beam image onto wafer or mask blank 1008 to form the final shaped beam image. The system shown FIG. 10, having a minimal number of components, allows shaped electron beam columns to be constructed utilizing a minimum amount of space and length m particular, thus reducing the effect of electron-electron interactions.

FIG. 11 shows a conventional variable shaped electron beam column, contrast to the electron beam column shown FIG. 10. An electron beam is formed at electron source 1101. Electron source 110] may be a thermionic cathode such as lanthanum hexaboride, LaB₆ , or a single gated photocathode similar to that shown m FIG. 3. The electron beam is shaped by square aperture 1102 to form a shaped electron beam. The shaped electron beam is focused by electron lens 1103 into region 1110. Spot shaping deflector 1104 deflects tne electron beam at focus region 1110 so that the shaped electron beam is shifted. The shaped electron beam is then passed through square aperture 1105 to form an intermediate shaped electron beam. Square aperture 1105 passes that portion of the electron beam that overlaps with the aperture and blocks that portion of the electron beam outside the aperture so that only a portion of the image formed by square aperture 1102 is passed into the intermediate shaped beam image. Demagnification lens 1106 demagnifies the image and focuses the image onto a final shaped beam image 1108 on a wafer or mask blank 1109.

The throughput of lithography systems is generally limited by the total beam current. As the total beam current is increased, electron-electron interactions m the beam cause blur, which causes a degradation of the resolution m a pattern written on a target. When the total beam current is distributed evenly among several columns, each column generates a lower beam current. Consequently, each column generates images having less blur and thus higher resolution. Further, distributing the beam current allows for higher electron current density, which allows for higher throughput when the columns are used m lithography.

In accordance with one embodiment of the present invention, multiple microcolumns 1201-1 to 1201-4 are arranged m an array 1200, such as depicted schematically m FIG. 12A. Any microcolumn that uses a combination^"of photocathode array 910 and gate electrodes 909 (FIG. 9) as a source of charged particle beams is suitable. For example, microcolumn 900 (FIG. 9) is an exemplary embodiment of any of microcolumns 1201-1 to 1201-4. In FIG. 12A, beam source 1202-1 refers to the combination of photocathode array 910 and gate electrodes 909. Beam sources 1202-2 to 1202-4 are similar to 1202-1. Microcolumns 1201-1 to 1201-4 of array 1200 selectively emit respective beams 1204-1 to 1204-4 onto target 1206, which is either a semiconductor wafer or a mask blank for electron beam lithography.

The pitch (center to center distance) between adjacent microcolumns m array 1200 in both dimensions of the array is, for example, 2 cm, where this is the typical diameter of the individual microcolumns including the housing of each microcolumn. The

"footprint", i.e., diameter, of a single microcolumn is, e.g., 2 cm or less, and therefore many microcolumns can be arranged an array which occupies the footprint (area) of a conventional 9 inch or 12 inch diameter semiconductor wafer, for example.

Of course the arrangement of microcolumns of array 1200 can be varied to be, e.g., any 2-sided array of size M by N, where M and N represent the number of microcolumns m the respective X and Y directions. FIG. 12B schematically depicts a plan view of the N by M array. This arrangement is not limiting.

By using the photocathode array 910 to generate multiple beamlets, the beam current m each microcolumn is divided into multiple beamlets, which further reduces the blur due to electron-electron interactions and therefore allows for a higher total beam current each microcolumn. Consequently, throughput of written patterns is increased. In one embodiment, the microcolumns of array 1200 use writing strategies such as the well known MEBES raster scan writing approach currently used for making masks with a single electron beam column. See e.g. IBM U.S. Patent No. 4,818,885 and Bell Labs U.S. Patent Nos . 4,668,083 and 3,900,737, incorporated herein by reference their entirety disclosing this MEBES writing approach. Thus a system architecture that uses the array 1200 of microcolumns includes, one embodiment, the basic data path and many other advanced techniques such as multi-path and multi-pixel techniques developed for MEBES. See IBM U.S. Patent No. 5,621,216 and Etec System Inc. U.S. Patent Nos. 5,393,987 and 5,103,101, which are incorporated herein by reference in their entirety. Also, other well known electron beam lithography techniques such as gray scale or shaped beam can also be used conjunction with array 1200; see e.g. IBM U.S. Patent Nos. 5,213,916, 5,334,467, 4,568,861, and 4,423,305, which are incorporated herein by reference in their entirety. The above described examples are demonstrative only. Variations that are obvious to one skilled m the art fall within the scope of this invention. As such, this application is limited only by the following claims.

Claims

CLAIMS We claim:

1. A charged particle beam column apparatus, the apparatus comprising: a support for a target; and an array of charged particle beam columns arranged to expose the target to charged particle beams, wherein each charged particle beam column comprises : a beam source that selectively generates a plurality of charged particle beams ; an anode coaxial with the charged particle beams that accelerates the plurality of charged particle beams from the beam source; and a lens coaxial with the charged particle beams that demagnifies the charged particle beams .

2. The apparatus of Claim 1, wherein the array is square shaped.

3. The apparatus of Claim 1, wherein the array comprises an array of N by M charged particle beam columns, where N is not equal to M.

4. The apparatus of Claim 1, wherein each charged particle beam column is a microcolumn.

5. The apparatus of Claim 1, wherein the beam source comprises a photocathode array that selectively supplies multiple electron beams when illuminated, the photocathode array comprising: at least one photocathode, the at least one photocathode having a gate electrode that modulates the electron beam in response to a voltage applied to the gate electrode; and at least one pad, the at least one pad being electrically connected to the gate electrode of the at least one photocathode, thereby allowing external control of the gate electrode.

6. The apparatus of Claim 5, wherein the gate electrode of the at least one photocathode includes at least one segment .

7. The apparatus of Claim 5, wherein the gate electrode of the at least one photocathode includes four segments.

8. The apparatus of Claim 5, wherein the gate electrode of the at least one photocathode includes eight segments.

9. A method of imaging a pattern onto a target, the method comprising the acts of: generating a plurality of charged particle beams; aggregratmg the beams into groups; selectively controlling each of the groups of charged particle beams; and directing each of the groups onto the target.

10. The method of Claim 9, wherein each of the groups is contained m one charged particle beam column.

11. The method of Claim 10, wherein the charged particle beam columns are arranged m an array.

12. The method of Claim 11, wherein the array is square shaped.

13. The method of Claim 11, wherein the array is rectangle shaped.