JP7164300B2 - Contamination identification device and method - Google Patents

Description

The present invention relates generally to apparatus and methods useful in cleaning surfaces. In particular, the present invention relates to apparatus and methods useful in cleaning the surfaces of components typically used in the semiconductor industry, optics, and the like. The disclosed apparatus and methods can be used to extend the useful life of photomask reticles.

[Description of related application]
This application is a continuation-in-part of U.S. patent application Ser. No. 14/656,206, which is a continuation of U.S. patent application Ser. No. 14/077,028, filed November 11, 2013 (now U.S. Pat. No. 8,741,067). Continuation application, this U.S. patent application being a continuation of U.S. patent application Ser. This U.S. patent application is a continuation of U.S. patent application Ser. No. 12/277,106, filed November 24, 2008 (now U.S. Pat. , is a continuation of U.S. patent application Ser. No. 12/055,178, filed Mar. 25, 2008 (now U.S. Pat. This application claims priority to US Provisional Application No. 60/954,989, filed on May 9, 2001, which patent documents are incorporated by reference and their disclosures in their entirety are hereby incorporated by reference.

Electromagnetic radiation has long been used for surface cleaning. Examples of these processes include removal of surface contamination, removal of thin material layer coatings, such as paint, or removal of oil from metal work surfaces. Some of the earliest examples utilized flashlamp radiation sources. These systems may be of limited application in view of the peak power achievable.

Lasers are increasingly being used in these types of processes because of the high peak powers achievable, high energy stability and wavelength selectivity. These features allow for high localization, improved material selectivity, and depth control of the cleaning effect. Laser surface cleaning processes can be roughly divided into surface contaminant layer removal and particle removal. Removal of the surface contaminant layer is usually accomplished by laser ablation. Particle removal requires the removal of contamination as a whole.

Cleaning processes belonging to both categories can benefit from the use of pulsed laser radiation to provide high peak power. In particular, short pulse radiation results in improved processing. Short pulse radiation has been found to reduce the heat affected zone in laser ablation processes. This allows for improved ablation removal as well as fine control of removal depth. Short-pulse radiation can also facilitate particle removal by increasing the rate of increase of heat within the particles and/or the substrate, thereby increasing the acceleration force that causes particle removal.

Substrate damage can be a challenge for both ablation and particle removal processes, and several techniques have been developed to minimize these effects. For the ablation process, selecting wavelengths that increase the absorption of contaminants can reduce fluence requirements and thus substrate damage. Additionally, using multiple pulses for total contaminant removal may reduce the required fluence. However, substrates with high absorption at selected wavelengths are more likely to be ablated with contamination even with wavelength selection and multi-pulse ablation processes. Termination possibilities of the removal process at the substrate interface would be limited in these cases. This problem is significantly increased for small size contamination, as the absorption cross-section of the contamination is reduced with respect to the substrate.

As with ablation removal processes, particle removal processes can also cause substrate damage for delicate substrates and substrates with high absorption at the processing wavelength. This problem is exacerbated with small particle removal due to increased adhesion between the particles and the substrate and self-focusing of the laser under the particles. For particle cleaning processes, developed apparatus and methods that reduce the risk of substrate damage require control of the environment above the contaminated surface. Examples of particle laser processes that can reduce fluence levels include wet laser cleaning, steam laser cleaning, and increased humidity cleaning. Combining laser cleaning processes with other cleaning processes (including etching, organic solvents, and ultrasound) has been found to increase cleaning effectiveness and can reduce the potential for substrate damage. However, with the exception of the dry laser cleaning process, all of the particle removal processes described require access to the environment above the substrate surface. This may be impractical for some systems.

Another dry laser particle cleaning process has been developed. Laser sonic cleaning and laser shock wave cleaning are also popular dry laser cleaning methods for particle cleaning. Laser sonic cleaning requires direct excitation of the substrate and is therefore subject to a high potential for substrate damage, especially for the small particles discussed. Laser shock wave cleaning has been found to improve particle removal and reduces the risk of particle damage by focusing the laser above the substrate surface and exploiting shock wave interaction with the particles. can be done. This technique also has increased difficulty when utilized for small particle removal. Additionally, the shock wave may damage other sensitive features on or near the surface of the substrate. This is especially true when there is sensitive material above the substrate surface, as it requires a relatively high laser intensity focused above the substrate to produce a shock wave.

Even with state-of-the-art dry laser technology, such technology may also be limited if access to the environment above the substrate is not feasible (eg, closed systems). The removal process simply moves the particles to another location on the substrate for closed systems, since the particles are removed from the surface as a whole. Typically, these techniques utilize additional controls and methods to completely remove particles from the substrate being cleaned. These methods include the use of directed air flow, reduced pressure (vacuum), or gravity, most of which require unrestricted access to the environment above the substrate surface.

Semiconductor manufacturing is one of the major industrial sectors that utilize surface cleaning processes, including laser cleaning methods. Many of the required cleaning processes have tight tolerances on acceptable substrate damage levels. Additionally, due to small product features, it is necessary to remove very small particles to avoid product damage. Cleaning is a challenge in many wafer processing steps, and such cleaning includes long-term contaminant layer (eg, resist removal) and particle contamination removal.

Surface cleaning is a requirement for optics (eg, photomasks) used in wafer fabrication processes. In particular, with respect to photomasks, contamination build-up is observed during normal use of the mask during wafer printing processes. These masks are exposed to deep ultraviolet (DUV) radiation during normal processing used in printing on wafers. Exposure to this radiation results in contamination growth in the form of small particles that absorb the illuminating radiation. This growth is commonly called haze.

Haze generation is a problem for wafer printing processes because as particles increase in size, they block much of the light being transmitted through the photomask. Ultimately, haze contamination absorbs enough light to cause defects in the printed image of the photomask on the wafer. The photomask surface must be cleaned before haze contamination reaches this level. This cleaning requirement has the effect of shortening the usable life of the photomask. This is because the processes currently used to remove haze degrade the absorbing film on the mask. For partially absorbing films, current cleaning methods reduce film thickness, thus adversely affecting film permeability and phase properties. Phase and/or transmission variations reduce reticle life by altering the geometry of printed features on the wafer beyond acceptable tolerances. In order to continue manufacturing once the usable life of the photomask is exceeded, duplicate photomasks must be made. Two sets of the same also need to be available while the contaminated photomask is being cleaned. There may be a time requirement of several days before cleaning and verifying the photomask, as the cleaning process is typically performed at a different facility. As the feature sizes required for semiconductor manufacturing decrease, the size of haze growth that can cause print defects also decreases. This increased sensitivity to haze growth means that modern photomasks must be cleaned frequently and their usable lifetimes are short.

Additionally, the composition and source of haze and most contaminants on semiconductor substrates are often difficult to locate. To this end, it would be beneficial to develop tools and methods that not only remove contaminants but also allow users to ascertain the physical and chemical properties of contaminants. This can be accomplished by collecting the contaminants for analysis as they are removed from the substrate. Contaminant collection and subsequent analysis help to indicate the source of the contaminants. This information can then be used to reduce or even eliminate contamination problems in the semiconductor manufacturing process, thereby saving both processing time and manufacturing costs.

Discerning the chemical or elemental composition of particles on the surface of a substrate is limited by the technology and resolution of the instrumentation used for analysis. Contaminants may be deposited as very small individual particles or thinly distributed films along surfaces, but many chemical analysis techniques do not have the resolution to detect or locate chemical composition. do not have. For example, on reticles used in photolithography that are exposed to other types of cleaning fluids of varying chemical composition, contaminants remaining after each cleaning may not be detectable chemically or by particle inspection. Over time and with constant exposure to both energy and ambient airborne molecular contamination, these remaining molecules tend to bond with each other and grow on the surface. Such molecules are on the order of nanometers, sometimes microns, and cannot be detected with standard chemical analysis tools. However, these molecules exist and grow to sizes observable by particle inspection instruments down to tens of nanometers. As these molecules grow, they interfere with wafer manufacturability, especially when they grow to defects large enough to be discernible on the wafer. In such cases, identifying the composition of the molecules on the reticle allows the manufacturer to trace the source.

The use of alternative cleaning methods to remove photomask surface haze contamination is hampered by the use of pellicles attached to the photomask surface. A pellicle consists of a frame glued to the photomask surface and a thin membrane or membranes stretched across the width of the pellicle frame. Pellicles are used to prevent externally generated particles from accumulating on the surface of the photomask from adversely affecting the printing process. Externally generated particles accumulate on the membrane above the mask surface where the impact of these particles on the printing process is significantly reduced. Except for a small filter valve on the pellicle frame to allow pressure balancing, the top surface of the photomask is effectively sealed from the local environment by pellicle mounting.

Currently accepted methods for haze removal require the wafer manufacturer to send the contaminated photomask back to the mask maker or a third party. At this point, the pellicle frame is removed from the photomask, the mask is cleaned, inspected for defects, and a new pellicle is attached to the photomask, often re-inspecting the mask for particle defects. and then send it back to the wafer manufacturer. This typically takes several days to complete, increases photomask cost due to additional processing, and degrades photomask quality due to the cleaning process. In addition, since adhesive from the pellicle is typically removed and falls onto the printable areas of the photomask, there is a small chance that the photomask will be unusably damaged by the haze removal process.

Current technological efforts to ameliorate the problems associated with haze growth on photomasks focus on processes that can be implemented prior to adding a pellicle because of the problems associated with thorough pellicle cleaning. These technological efforts are primarily focused on the use of alternative chemicals in surface pretreatment and cleaning processes. It has been found that the latter alters the haze-contaminating species (active species) but does not stop their growth. Both regions show, at best, a reduction in growth rate, but do not eliminate the requirement for cleaning. It has recently been found that the use of an inert environment reduces the growth rate of haze formation on photomasks. Utilization of this method requires full control of the photomask exposure environment, including all process equipment. As with other methods under development, this process has the potential to reduce growth rate, but does not eliminate the requirement for cleaning and the detrimental effects of cleaning.

In one aspect of the invention, a method of cleaning a surface of a substrate is provided. The method includes directing a laser toward a substrate with contaminant particles attached thereto, causing an increase in temperature of the substrate, and transferring thermal energy from the substrate to the particles to decompose the particles. The laser has a wavelength that is substantially the same as the local maximum of the absorption spectrum of the substrate.

In another aspect of the invention, a method is provided for extending the usable life of a photomask substrate. The method includes the steps of directing electromagnetic radiation through a protective material toward a substrate to which contaminant particles are attached, causing a temperature rise in the substrate, and transferring thermal energy from the substrate to the contaminants to break up the particles. including the step of The electromagnetic wave has substantially the same wavelength as the local maximum of the absorption spectrum of the substrate.

In another aspect of the invention, a method is provided for cleaning a surface of a substrate at least partially contained within a pellicle. The method includes the steps of directing a laser through a pellicle film toward a substrate having a layer of contaminant particles deposited thereon, producing a temperature rise in the substrate, and transferring thermal energy from the substrate to the layer of particles to disperse the particles. Decomposing at least a portion of the layer. The laser has a wavelength that is substantially the same as the local maximum of the absorption spectrum of the substrate.

In another aspect of the invention, a non-destructive method for surface contamination release and analysis is provided. The method can be performed on substrates under atmospheric pressure conditions, and the method allows in situ detection and analysis of substrates with one or more vacuum-compatible portions. The focus of this method is to liberate molecules from the substrate to produce airborne molecular contamination (AMC), which is then detected or analyzed with high sensitivity. The method includes a source of electromagnetic radiation that strikes the substrate and liberates surface contamination. The liberated AMC is then sent to an instrumentation system for detection or analysis. Alternatively, loose contamination may be deposited on a secondary collection substrate, the purpose of which is to concentrate the contamination for internal analysis or delivery to an external metrology system. A partial local vacuum may be used to help collect loose molecules or particles from the surface.

Thus, for the purpose of enabling a better understanding of the detailed description of the invention herein, and of the contribution of the invention to the art, the present disclosure is omitted. Certain embodiments of the invention have been outlined rather broadly. There are, of course, additional embodiments of the invention that will be described below and which will form the subject matter of the claims appended hereto.

In this regard, before describing at least one embodiment of the invention in detail, the invention is limited in its application to the details of construction and arrangement of components set forth in the following description or illustrated in the drawings. It should be understood that there is no In addition to the embodiments described, the invention is capable of other forms and of being embodied and carried out in various ways. Also, it is to be understood that the phraseology and terminology employed in the specification and abstract are for the purpose of description and should not be construed as limiting the invention.

Accordingly, those skilled in the art may readily utilize the concepts underlying this disclosure as a basis for designing other structures, methods and systems for carrying out the several objects of the present invention. be understood. It is important, therefore, that the invention as set forth in the claims be construed to include such equivalent constructions insofar as the claimed subject matter does not depart from the spirit and scope of the invention.

A number of figures are attached showing various embodiments of the present invention.

Fig. 4 is a schematic diagram showing the state of laser excitation and surface contamination; 1 is a schematic illustration of a substrate surface showing how to decontaminate, according to certain embodiments of the invention, a number of species can be removed from the mask, and these species can be in the form of gases, liquids, solids, etc.; It is a figure which shows that it may be in. 1 is a schematic representation of a photomask surface with a thin film absorber on top containing contamination on the film and substrate. FIG. 2 is a graphical representation showing plots of quartz absorption spectra from the far ultraviolet region to the far infrared region of the electromagnetic spectrum; FIG. 2B is a schematic representation of a photomask surface with a thin film absorber comprising a pellicle attached, showing the possible presence of contamination on the film and/or substrate, according to certain embodiments of the present invention; . FIG. 2B is a schematic illustration of a photomask with a pellicle showing a laser beam focused through the pellicle onto a surface; Fig. 4 is a schematic diagram comparing the beam spot size on the pellicle and the beam spot size on the mask caused by focusing; FIG. 1B is a schematic illustration of a photomask with a pellicle showing a laser beam focused through the pellicle onto a surface; FIG. FIG. 4B is a cross-sectional view of a Gaussian distribution of beam energy and a corresponding resulting temperature distribution; FIG. 4 is a cross-sectional view of a top-hat beam energy distribution and corresponding resulting temperature distribution, wherein Gaussian, flat-top, and/or top-hat energy distributions can be used, according to certain embodiments of the present invention; It is a figure which shows the state which can be performed. 1 is a schematic representation of a photomask with a cold plate in contact with the bottom of the mask, and according to certain embodiments of the invention, the contact point is, for example, a flow of water (or other liquid or gas) through the cold plate; Fig. 3 shows that it may be either an electrical contact for thermoelectric cooling; FIG. 4B is a schematic diagram illustrating a technique for forced air cooling of an area on a photomask, where the airflow is directed at a pellicle frame, according to one particular embodiment of the present invention; 1 is a schematic diagram showing a single pass of a laser beam across a surface to minimize localized thermal deposition, showing a single row or column with large lateral spacing between spots; is a diagram. FIG. 2 is a schematic diagram showing two passes of a laser beam across a surface for minimizing localized thermal deposition, with two sets of beam spots overlapping each other with a large spacing between sets of pulses; FIG. Fig. 3 shows a single row; Fig. 3 is a schematic diagram showing multiple laser passes over an area of a substrate to achieve complete cleaning of a section of the substrate; Fig. 3 is a schematic diagram showing a second dimension of surface cleaning; Fig. 3 is a schematic representation of the use of non-sequential pulses on a surface; Fig. 10 is a schematic diagram showing the use of a laser pulse pattern to control the location of residual material; Fig. 3 is a schematic diagram showing the use of gravity to control the location of residual material; Fig. 3 is a schematic diagram showing the use of gravity to control the location of residual material; 1 is a schematic illustration of a contaminated substrate surface provided with a thermocouple or infrared temperature monitoring device; 1 is a schematic representation of a contaminated substrate surface provided with an imaging system, a microscopy system, a spectroscopy system, or a combination system for contamination analysis; FIG. 4 is a schematic representation of a contaminated substrate surface provided with an imaging system, showing the imaging system and the laser beam delivery being in a common path; Fig. 3 is a system schematic showing robot loading and X/Y/Z stage motion of a substrate with respect to a laser beam; 1 is a box diagram of an exemplary wafer fabrication process utilizing a photomask wet cleaning process; FIG. FIG. 4 is a box diagram of a wafer fabrication process flow including the use of laser photomask cleaning without pellicle removal; FIG. 3 is a box diagram of a wafer fabrication process flow including the use of laser photomask cleaning in which no additional mask set is used during the cleaning process; Fig. 3 is a schematic diagram illustrating how the identification is performed;

The invention will now be described with reference to the figures contained in the drawings, throughout which like reference numerals refer to like parts. Certain embodiments of the present invention provide a laser surface cleaning method with reduced risk of substrate damage.

FIG. 1A shows that excitation energy 2 comes from an energy source, e.g. or by conduction). However, energy sources other than lasers can also be used (e.g., lamps and other devices capable of emitting energy throughout the electromagnetic spectrum, such as X-rays, microwaves, infrared, generators such as near-ultraviolet). Also, the surface may be of any material (eg, the surface of a silicon wafer). The resulting temperature rise upon contamination typically causes removal by the action of heat, the effect of which is shown in FIG. Degradants 5 include, but are not limited to. Additionally, contaminant particles 3 are found on the photomask as shown in FIG. 2, which shows contaminant particles 3 on the substrate 4 and thin film absorber 7. FIG.

According to certain embodiments of the present invention, the method reduces the risk of substrate damage, as temperatures typically used to perform surface cleaning are below the thermal damage level of the substrate 4 material. The risk of substrate damage is also generally mitigated by other techniques, as such techniques can sometimes utilize relatively long pulse widths, thereby reducing the potential for multiphoton absorption processes. is often reduced.

The exemplary methods described above can generally improve the removal of small contaminants/particles because such methods are minimally dependent on particle size. The method may be particularly advantageous for applications in which the environment above the contaminated substrate is substantially or completely enclosed. In these cases, the method may further include directing the beam through a material provided with respect to the substrate and forming part of the substrate's environmental enclosure. For example, the method of the present invention can be used to clean haze contamination from the surface of a pellicled photomask.

Decomposition of contaminant species has been suggested to be beneficial in laser surface cleaning processes. However, prior to the development of embodiments of the present invention, there has been no disclosure of a process for thermal surface cleaning using laser heating of a substrate.

According to a particular embodiment of the invention, the method comprises the steps of selecting a laser wavelength substantially matching the strong absorption of the substrate and setting the laser energy and pulse width to produce the desired cleaning effect. including. Increased absorptivity of the substrate may allow lower laser energy for the cleaning process in some cases. Therefore, the risk of damage to adjacent materials that may interact with the laser when it is directed at or reflected from the surface is reduced. Although not a requirement, according to certain embodiments of the present invention, wavelengths are selected that are also highly absorbed by one or more contaminants, since this produces the desired thermal removal effect. because it improves The use of multiple laser wavelengths and/or laser energies may be used in the cleaning process, especially when the substrate is composed of more than one material. Multiple laser energies may also be used on the same component, typically when the same component undergoes one or more material property changes during the first step of the desired cleaning process. Multiple wavelengths can be produced, for example, by utilizing multiple laser sources or a single tunable laser source or both. Multiple energies may be used by controlling the output energy of the laser source using controls that are internal or external to the laser.

Example

The following is an example of a method according to an embodiment of the invention utilized for surface cleaning of haze contamination from photomask substrates used in wafer fabrication processes. This example can be used throughout additional embodiments of the methods of the invention described.

Certain embodiments of the present invention can be used for surface cleaning of wafers (eg, silicon wafers). Some form of haze growth is also observed on these substrates and such haze growth can sometimes be problematic if it is not removed prior to wafer printing. The use of environmental controls has been suggested for controlling haze growth on silicon wafers. However, one particular embodiment of the invention is for haze mitigation or other forms of decontamination on surfaces, such as silicon wafers. Specifically, according to certain embodiments of the present invention, haze can be removed by laser exciting a silicon wafer substrate below the thermal damage threshold.

Methods of certain embodiments of the present invention can wash away contaminant precursors because such methods typically do not utilize direct absorption of substances located on the substrate. is. Thus, the method of certain embodiments of the present invention can act as a surface pretreatment technique to reduce contamination generation rates. For example, to reduce photomask haze growth, the method of the present invention is applied prior to use in a wafer fabrication process and haze growth precursors (e.g., acid residues, water, etc.) or nucleation sites are removed. Or better to rearrange. Other techniques can be used in conjunction with the present invention to further suppress haze regrowth or reformation after a reticle has been treated with the present invention. For example, surface pretreatment prior to pellicleing or environmental control before, during, or after processing can extend reticle life by reducing haze regeneration or reformation rates.

Application of the methods of certain embodiments of the present invention to substrates composed of multiple materials may require consideration of beam parameters, including material parameters as well as excitation wavelength selection. Based on the cleaning process, it is particularly desirable that all contaminated areas of the substrate reach temperatures substantially close to those required for removal, typically without exceeding the thermal damage threshold of the substrate. Typically, the laser energy required to bring one of the materials to the processing temperature can cause thermal damage to the other material, especially if there is a substantial difference between the materials' absorptances. is. The local fluence of the beam can be controlled based on the material being exposed.

According to certain embodiments of the present invention, long laser pulse durations, including maximum continuous wave (CW) lasers and such continuous wave lasers, are used to improve thermal equilibrium between materials having significantly different absorption coefficients. . However, the use of such long laser pulse widths results in the highest thermal gains in the system, so if the material located adjacent to the substrate surface has a thermal damage threshold below the process temperature, long laser pulse widths may be used. may not be used.

According to one particular embodiment of the invention, a laser wavelength is chosen that exhibits significant absorption in all of the materials on the substrate. The same laser energy can then be used to produce a desired process temperature, eg, below any of the damage thresholds of the substrate material. By considering thermal properties (including diffusivities), it is possible to exploit heat transfer between different materials. This allows the use of reduced process fluences to achieve ablation for the entire substrate in some cases, especially when the thermal energy flow from the highly absorptive material is preferred over the less absorptive material. be.

Beam parameter control is particularly desirable in embodiments of the invention relating to photomask surface cleaning of haze contamination. Wavelength selection is highly desirable, for example, because of the physical structure of typical photomasks. Referring to FIG. 2, a photomask typically consists of a quartz substrate 4 with a thin absorbing film 7 applied over its critical surface. For metal films, there will typically be a sizable absorption coefficient for most of the laser wavelengths that can be produced. However, in the case of quartz substrates, generally speaking, there is a limited wavelength range over which the substrate has substantial absorption and laser sources are generally available. Given the relationship between the thermal properties of quartz and the thermal properties of metal layers, one particular process embodiment of the present invention utilizes wavelengths that are highly absorbed by the quartz substrate material, the utilization of which is This is because mutual heat transfer occurs preferentially from the quartz to the metal layer.

The above discussion also applies generally to the case of partially absorbing photomask films. In general, thermal flow for photomasks with partially absorptive films occurs preferentially from quartz to the film, because such films typically contain metallic components and the thermal diffusivity of quartz is This is because it is relatively low. However, there may be wavelength regions where these films do not absorb as much as they do in pure metal films. This has the potential to increase the wavelength range that can be used to preferentially excite the quartz section of the substrate. In addition to thermal flow, thermally induced material changes (eg, oxidation, annealing, etc.) must be considered for partially absorbing films. The phase and transmission properties of these materials are critical to their function and can be altered by heat treatment. It may be necessary to limit the maximum temperature of the removal process if thermal material changes adversely affect membrane performance. Controlling process uniformity by pulse shaping or pulse overlapping may be necessary when thermally induced material changes provide beneficial effects.

A specific example of a representative method of the present invention is the removal of haze by ammonium sulfate from the surface of a photomask. Temperature and Other Areas - Ammonium sulfate is expected to decompose at temperatures above 280°C. The lowest thermal damage point for a typical photomask will typically be the melting/reflow point (ie, about 1600° C.) for the base quartz substrate. Therefore, there are potential processes that can produce temperatures for decontamination/cleaning that are below the damage level of the substrate material.

It is important to note that the species being removed from the photomask itself typically only determines the process temperature requirements. Although it may be advantageous to provide the contaminants with a significant absorption rate, this is not a requirement. As noted above, the relative absorption rates of substrate materials are generally considered because of potential differences in material absorption properties. In particular, it may be desirable for quartz substrates to have substantial absorptance at process wavelengths, primarily because of the preferential thermal flow to the absorptive film.

Quartz substrates used in photomasks are typically specially designed to have high transmission in the deep ultraviolet (DUV) wavelength range, as shown in the quartz absorption spectrum of FIG. This is typically accomplished by using synthetically formed substrates with extremely low impurity levels. Except for relatively low absorption near 3 μm wavelengths, these materials typically also have high transmission in the infrared region. The main absorption of these substrates generally occurs at wavelengths either below 0.2 μm or above 8 μm. Short wavelengths are not within a particularly desirable wavelength range, because such short wavelengths are usually significantly absorbed by air, and short wavelengths have high photon energies, and This is because there is a high possibility of causing a multiphoton process.

Selecting wavelengths above 8 μm, for example close to the 9 μm quartz absorption, is particularly desirable according to certain embodiments of the present invention. This typically results in high absorption in the quartz substrate without high environmental absorption. This wavelength has an additional advantage, especially if the photomask has a metal film layer (e.g., chromium), because the reflectivity of the metal film increases with increasing wavelength in this region. . This typically reduces the light that must be absorbed by the film and generally improves the thermal excitation bias for quartz. This wavelength can also provide advantages for photomasks with partially absorbing film coatings (ie, MoSi) due to the relatively high absorption coefficient for quartz. In general, the film material temperature reached at constant fluence should be about the same as that for quartz due to the high quartz absorption and high thermal diffusivity of partially absorbing films compared to quartz. This is expected to be true even if the partially absorbing film has a relatively high absorption coefficient within this wavelength range.

The process just described for use with certain embodiments of the present invention typically renders photomasks usable by replacing current cleaning processes used to remove haze from photomask surfaces. Extend life. Unlike typical chemical cleaning processes used for haze cleaning, the laser ablation process as certain embodiments of the present invention generally reduces the absorption film thickness and/or linewidth of the absorbing film. does not reduce This is the limit to the number of conventional "cleaning processes" that can be performed before the photomask is no longer usable, since material loss is a result of these processes. This is especially true for photomasks with partially absorbing films, since material loss results in phase loss and increased transmission through the film. By design, the performance of absorptive film photomasks is highly dependent on film phase and transmittance. It is contemplated that an unlimited number of cleaning cycles can be used in the laser cleaning process according to the present invention.

It has been found that employing temperatures below the critical range can lead to material changes in the partially absorbent membrane. For example, a partially absorbing MoSi film may be annealed at a first temperature, and the effect of annealing the film is to reduce the phase retardation or significantly reduce the transmittance of light transmitted through the film, and thus the cleaning process. It may be necessary to operate below the first temperature. If not, the life of the partially absorbing film photomask will be shortened if the photomask is subjected to the conventional wet "cleaning process" currently used for haze removal. However, the temperature of the process of the present invention can be adjusted by adjusting the energy provided to the surface (e.g., controlling pulse duration, pulse amplitude, CW energy, etc.), thereby avoiding film critical temperature tolerances. Can be fine-tuned.

However, if the effect of annealing the film is to increase the phase retardation of light transmitted through the film with minimal or no loss in transmission, then the cleaning process should be performed above the annealing temperature. may be advantageous. Standard wet "cleaning processes" are essential in the manufacture of photomasks and can produce unacceptably low phase retardation for partially absorbing films even prior to use. Additionally, wet cleaning processes may be required in addition to use of the present invention. For example, a wet cleaning process may be required if defects not related to haze are present or appear on the photomask. Inducing a material change in the partially absorbing film during the cleaning process of the present invention may extend photomask life by recovering the loss of phase retardation due to the wet cleaning process. Also, an example of thermal modification of the partially absorbing film using the method of the present invention extends the life of the photomask reticle by recovering the phase retardation loss during the wet cleaning process itself (without the requirement for haze cleaning). It is also conceivable that the

One of the reasons for using a vigorous wet cleaning process is that the removal of the pellicle frame from the photomask leaves behind an adhesive residue. Wet cleaning processes generally have a detrimental effect on adhesive residue, thereby contaminating the working area of the mask with the adhesive, because wet cleaning processes are generally difficult to localize. However, some of the laser cleaning processes disclosed herein can be localized from adhesive residue, thereby leaving the adhesive residue unaffected. Controlled removal of the pellicle frame and bulk of the adhesive followed by a laser cleaning process as an embodiment of the present invention allows subsequent pellicle installation without wet cleaning requirements (strong or otherwise) is. This is especially true when the laser cleaning process as an embodiment of the present invention utilizes alternative bonding methods or is combined with the use of multi-part pellicles that do not require adhesive exposure for pellicle exchange.

A method according to one particular embodiment of the present invention can be used for photomask haze cleaning, and such method does not require pellicle removal. These laser cleaning methods are typically performed through the pellicle membrane material without adversely affecting the pellicle membrane properties, which are shown in FIG. 9, and substrate pellicle adhesive 10 are shown.

In this case, it is customary to consider the absorption of the pellicle film 8 at the process wavelength and the energy density (fluence) at the surface of the substrate 4 . As with the substrate 4 and substrate film coating 7, the cleaning process generally does not cause the temperature of the pellicle film to rise above the damage threshold. However, depending on the pellicle film, there may be a significant absorption in the pellicle film near the 9 μm absorption peak for the quartz substrate. However, operation in the region of significant pellicle membrane absorption is still possible, since the pellicle membrane is positioned above the substrate surface.

FIG. 5A illustrates a technique for focusing the excitation energy 2 through a focusing lens 11 that produces a focused beam 12 that passes through the pellicle membrane and is deposited onto the surface of the substrate 4 . substrate film coating 7, thereby removing contaminant particles 3. The wavelength and focusing properties allow focusing at different heights and reduce the relative temperature rise of the pellicle membrane 8 . The temperature rise of any material is proportional to the fluence applied to the surface as shown in the equation:
ΔT∼F Equation 1
where ΔT is the temperature change in the material and F is the absorbed laser fluence.

For constant density or beam pulse energy, the fluence is inversely proportional to the square of the beam spot radius as given by
F~E/r2 Equation 2
where F is the fluence, E is the energy, and r is the radius of the beam on the substrate surface.

FIG. 5B shows the spot beam size on the pellicle. The ratio of the beam radius at the pellicle (pellicle beam 14) to the beam radius on the mask surface 4 (mask beam 13) is typically increased by focusing the beam through the pellicle and thus the photomask. The relative fluence on the pellicle frame compared to the substrate surface can be reduced. FIG. 5C is a side view showing the relationship between the convergence at surface 4 at the point of mask beam 13 and the unfocused energy at the point of entry of pellicle 8 (pellicle beam 14).

In addition to wavelength considerations, utilizing processes that produce large temperature rises in the system (eg, long pulse lengths or high repetition rates) may be limited by the damage threshold of the pellicle film. This is typically below the process temperature requirements for many photomask haze components.

pulse shaping

The pulse width, temporal pulse shape and spatial distribution of the laser can be used to enhance the cleaning process or extend the safe operating range for the treatment of certain embodiments of the present invention. Short pulse widths can be used to minimize the overall heat input to the system (substrate and contamination). A long pulse width can be used to maintain the process temperature for an extended period of time, thereby facilitating completion of the thermal removal process. Temporal pulse shapes can be used to control the temperature rise within the contaminant species. Long temperature rises can be used to produce an initial effect (eg, melting) followed by a secondary effect (eg, decomposition). A short rise time can facilitate evaporation of contaminants while limiting the decomposition process in some cases. Also, a combination of short and long temporal pulse shapes can be used to optimize the ablation process. The use of multiple pulses allows lower beam energies to be desired for complete cleaning, thereby further reducing the risk of substrate damage.

The spatial distribution of the laser beam can be used to expand the process window. For example, FIG. 6A shows a representative Gaussian spatial distribution 15 that can create a temperature gradient across the substrate 16, and FIG. A spatial distribution 17 is shown. Spatial distribution can be used to enlarge the process window. For example, employing a flat-top or top-hat spatial distribution allows uniform temperature rise within the beam spot, while a Gaussian distribution typically produces a temperature gradient within the beam spot. To avoid potential substrate damage, the maximum energy in the beam is typically limited by a Gaussian peak. As noted above, when more than one material is present on the substrate, long pulse widths can be used to achieve thermal equilibrium between the materials on the substrate.

thermal management

Since certain embodiments of the present invention involve heat-based processes, it may be desirable to manage the overall temperature of the system to avoid damage to heat-sensitive or easily contaminated materials. be. This is especially true for photomask haze cleaning that does not remove the pellicle. Pellicle membranes typically have a low thermal damage threshold. Therefore, it may be useful to avoid global system temperature build-up that can transfer to and/or damage the pellicle material. This includes the pellicle frame and the enclosed environment between the mask surface and the pellicle membrane.

Managing system temperature can be accomplished in several ways. The examples below show a few representative sample cooling methods, and it should be appreciated that others may exist. One approach to managing system temperature is through contact cooling. For example, the photomask may be placed in contact with a plate 17 which acts as a heat sink to draw heat generated on the front surface of the mask toward the back of the mask as shown in FIG. , with heat exchange tubes 18 and 19 . This reduces heat transfer to the environment above the mask surface, the pellicle membrane and the adhesive between the pellicle frame and the mask surface. Cooling can be accomplished in a variety of ways, including flowing water or other cooling fluids or gases over the mask and/or pellicle, thermoelectric cooling or laser-induced cooling of part or all of the mask and/or pellicle. cooling.

Another potential approach to temperature control is through forced convection cooling. The filtered and/or cooled gas or liquid stream is typically directed over portions of the mask, over the pellicle membrane, over the frame and/or over the adhesive, thereby It directly reduces the temperature build-up in these materials shown in FIG. A top flow 20, side flow 21 or bottom flow 22 of coolant can be used to control the temperature. This typically reduces the risk of pellicle membrane damage, as well as the risk of generating contaminant outgassing from the pellicle frame and pellicle membrane adhesive. In addition to hardware control of system heat build-up, heat build-up can be reduced by allowing increased processing time. The injected heat can be removed by applying a slow pulse rate to the system or allowing a delay between application of a series of pulses, in which case the overall system temperature will not rise above a critical level. do not have.

Pulse to pulse heat buildup can also be advantageously controlled, and such pulse to pulse heat buildup may depend on the thermal properties of contaminants, substrates and/or adjacent materials. In general, pulse-to-pulse heat build-up can be controlled by reducing the number of laser pulses striking the surface per unit time. This temperature increase can also be controlled by increasing the distance between adjacent laser pulses. It may be particularly desirable to provide large lateral displacements between adjacent pulses, in which case materials (eg, pellicle membrane materials) are particularly sensitive to inter-pulse heat build-up. In this case, the process typically involves multiple hits of the laser beam at approximately the same location to effect thorough cleaning of the target surface. For example, a first series of laser pulses 13 impinge on the surface 4 with relatively large lateral spacing as shown in FIG. 9A. A second pass over the same area applies an additional series of laser pulses 13, which are slightly offset with respect to the first set of spots as shown in FIG. 9B. This process continues until the entire area has been exposed to laser pulses 13 as shown in FIG. 9B. Overlap in a second direction can be used according to certain embodiments of the invention to fully expose the substrate surface 4 as shown in FIG. 9D. According to certain embodiments of the present invention, the entire process is repeated and/or the overlap between passes is increased, especially when it is desirable for the cleaning process to have multiple pulses for complete removal. Varying the position of the beam relative to the surface as shown can be accomplished by moving the beam and/or moving the substrate. Additionally, by applying the pulses across the mask in a systematically distributed manner, the potential for heat build-up on the mask can be further reduced as shown in FIG. 9E.

Residue control

According to certain embodiments of the present invention, laser cleaning methods may produce residual material on the photomask surface depending on the decomposition products of the contamination. Even if the residues no longer adversely affect the use of the substrate material (i.e. even if the substrate has been effectively cleaned), there are still reasons to control their location or concentration. . Traditional methods of controlling residue formation, such as directing a stream of air, a stream of water, or providing reduced pressure to the substrate to be cleaned can be used according to certain embodiments of the present invention. However, for closed systems, such as pellicled photomasks, it is typically undesirable to use these environmental control factors. Therefore, another method of controlling the location of residual material is used for closed systems according to certain embodiments of the present invention. For example, according to certain embodiments of the present invention, the pattern of laser pulses is controlled. For example, FIG. 10 shows that the laser pulse 13 begins in the center of the surface of the substrate 4 and is directed in a pattern 23 of circles of gradually increasing diameter or squares of increasing size, leaving residual material preferentially shown in FIG. Fig. 4 shows an embodiment moved towards the edge of the substrate as shown; Another method according to certain embodiments of the present invention for controlling debris is to use gravity. The manner in which the photomask is placed face down is shown in FIG. 11A or in a slanted position 24 as shown in FIG. Allows for preferential deposition. Alternatively, the reticle is rotated (ie, spun at high speed) in conjunction with certain embodiments of the present invention to move residual material away from the mask center and/or into non-active areas on the reticle. In addition, reducing the temperature of regions of the photomask, pellicle, and pellicle frame in accordance with certain embodiments of the present invention results in preferential deposition of residual material on these surfaces, since this This is because the substance is likely to undergo a phase transition from the gas phase to the liquid phase or the solid phase as shown in FIG. For example, these cooling methods include, but are not limited to, flowing water, other fluids or gases, thermoelectric cooling, or laser-induced cooling in and/or around the preferred deposition regions.

Reduced haze growth and regeneration

The present invention can be used in conjunction with surface preparation or environmental control techniques to extend reticle life. Some of these techniques require processing prior to pellicle attachment, while others can be performed after pellicle formation. For example, utilizing a surface pretreatment method in conjunction with the present invention prior to pellicleization can increase the time between washes. This can be important when there is a limited number of times the method of the invention can be cleaned before additional cleaning (eg, wet cleaning) is required. One embodiment of the method of the present invention is to place a seed crystal or other nucleating material under the pellicle in the non-active area of the reticle. These seed crystals may act as preferential growth sites for haze. This can effectively reduce the concentration of residue and precursors available in the active regions of the photomask and reduce the growth rate in these regions. Another embodiment of this invention is to coat the surface of the photomask with a material that reacts with and/or neutralizes the residue and precursors liberated by the cleaning process of this invention. This can also reduce the haze growth rate of the active area on the mask by limiting the reactive species available.

Post-pellicle techniques can also be used in conjunction with the present invention. For example, environmental controls or manipulations both inside and outside the pellicle can be used in combination with the cleaning process of the present invention. One embodiment includes replacing the environment under the pellicle with a non-reactive gas after the cleaning process. This can be done without pellicle removal by gas exchange through filtered vents in the pellicle frame. Additionally, it may be advantageous in connection with the present invention to maintain an inert environment outside the pellicle to reduce haze regrowth or reformation. These combined processes may be important if the time between cleaning processes according to the method of the present invention can be extended and a limited number of cleaning processes can be used.

Additional post-pellicle environmental control embodiments of the present invention include evacuating the environment beneath the pellicle and introducing or introducing substances that react with or neutralize this environment with haze residues and/or precursors. exchange with This process can be performed before the cleaning process, during the cleaning process or after the cleaning process. In all cases, haze residues and/or precursor species liberated during the cleaning process react with introduced/exchanged materials to create non-haze forming species.

Additional post-pellicle techniques can be used in conjunction with the present invention to mitigate haze regrowth or reformation. These techniques can utilize the thermal effects of the present invention to alter surface morphology and/or surface composition to inhibit haze growth and reformation. For example, operation at or near the quartz reflow temperature can produce changes in the material state or morphological features of the quartz substrate. This can reduce or eliminate activation sites that are believed to cause nucleation of crystal haze growth, thereby reducing haze growth or reformation rates. Alternate embodiments combine surface pretreatment or environmental control methods with the thermal effects of the methods of the present invention to modify or eliminate activation/nucleation sites. The precursors can be activated by heat treatment or can interact with the surface under thermal excitation to reduce haze growth and reformation.

metrological technology

The method of certain embodiments of the present invention can also be used in combination with metrology techniques to monitor critical process parameters and/or assess the progress or completion of the cleaning process. Measurement of the locally generated temperature of the substrate material can be used, for example, in combination with cleaning processes. Temperature measurements may be evaluated prior to application of the cleaning process to verify potential temperature-related damage. In addition, these temperatures can be monitored during the cleaning process to verify process control and/or mitigate the potential for material damage. For example, according to certain embodiments of the present invention, the temperature of the substrate and/or the absorbing film is monitored during the process, and such temperature provides feedback control of the energy applied to maintain the desired process or increases the temperature. It has the ability to turn off the cleaning process if too much temperature rise is detected. A number of temperature monitoring methods exist as shown in FIG. 12, including contact, eg thermocouple 26 and non-contact, eg infrared camera 25 techniques.

Metrology techniques are also used according to certain embodiments of the invention to analyze contamination for material properties before, during and/or after the removal process illustrated in FIG. Using contaminant identification prior to running the cleaning process, ideal processing parameters can be set. This allows the use of minimal process temperatures, thereby reducing the risk of substrate damage. This in-process contamination monitor can also be used to assess the completion of the cleaning process based on the intensity of the measured signal as the cleaning process progresses. Additionally, a monitor for substitute substances in the process can also be used to signal when the process is causing additional contamination and/or undesirable changes in the substrate material. can be done. This information can be used to control the process and/or reduce the risk of substrate damage and/or poor cleaning results.

There are many methods available for determining the chemical or elemental composition of particles on a surface. Such methods include, but are not limited to, spectroscopy, mass spectrometry, electrochemical analysis, thermal analysis, separation, hybrid techniques (including two or more techniques), microscopy. Embodiments of the apparatus described herein use spectroscopy, mass spectroscopy, or hybrids thereof. Reference herein to metrology techniques should be understood to broadly encompass various subsets of such techniques. For example, techniques encompassed within spectroscopy are Fourier transform infrared spectroscopy (FTIR), infrared spectroscopy, Raman spectroscopy, laser induced breakdown spectroscopy, electron paramagnetic resonance spectroscopy (EPR), nuclear magnetic resonance spectroscopy (NMR), AUGER electron spectroscopy, X-ray photoelectron spectroscopy, energy dispersive X-ray spectroscopy (EDS or EDX), electron energy loss spectroscopy (EELS), attenuated total reflection (ATR) and fluorescence spectroscopy Law. Commonly used techniques in photolithography to determine the chemical composition of particles on a substrate are EDX, time-of-flight secondary ion mass spectrometry (TOF-SIMS), Raman spectroscopy, FTIR, and liquid chromatography-mass spectrometry. method (LC-MS).

Liquid chromatography is a separation process that uses an absorbent material to separate particles from a solvent containing sample mixture. The solvent and sample mixture are dispersed across the adsorbent material at rates that differ from each other depending on the compound. This allows the compounds to be sequestered in different regions of the adsorbent. Liquid chromatography can be used in conjunction with mass spectrometry to more accurately identify molecules.

Mass spectrometry is a chemical analysis technique that can determine precise chemical composition by ionizing and mass sorting particles to produce a spectrum based on mass and ion charge. Mass spectrometry is destructive. Common techniques include quadrupole mass filters and TOS-SIMS.

A quadrupole mass filter (QMF) uses an ion source to ionize the particles and then an ion accelerator to accelerate the ions into a beam. The ion path is centered on four parallel rods, the ion path extends along these four parallel rods, and these four parallel rods generate a pulsed electromagnetic wave. used to make The rods act as two pairs 180° out of phase. This causes the ions to be radially displaced in oscillation along the beam as they move axially along the rod. Ions are not accelerated axially along the beam. A detector is provided at the end of the rod to locate the radial position of the ions. The radial position of ions on the detector is based on their mass-to-charge ratio as they pass through the quadrupole. One advantage of QMF is that it is relatively inexpensive compared to other mass spectrometry tools. There are a few drawbacks to using quadrupole mass filters. Quadrupole mass filters have limited resolution and must tune the peak and mass response, making these quadrupole mass filters unsuitable for pulsed ionization.

Time-of-flight secondary ion mass spectrometry (TOF-SIMS) is a technique in which a primary beam of ions impinges on the surface of a substrate. Secondary ions are liberated and collected by a detector. Based on the collision time from the primary ion beam and the detection time of secondary ions, the mass spectra of compounds present can be identified. This is the most sensitive method currently available with a resolution expressed in parts per billion (ppb). However, there are some problems. TOF-SIMS must be performed in vacuum to avoid contamination from the ambient atmosphere. This method is also destructive, the liberated ions are often part of the parent material of the substrate and not the very particles present on top of the substrate. For some applications, such as photolithography, this can damage the pattern applied to the reticle, rendering the reticle useless. This method is often qualitative and often results may vary depending on the operator of the instrument.

Raman spectroscopy uses monochromatic light (one wavelength) to bombard a molecule with photons. Energy is exchanged between the incident photon and the molecule, resulting in a change in the energy and wavelength of the outgoing photon. This phenomenon is called scattering. Different molecules undergo different energy conversions with photons, resulting in different wavelengths and spectra that are used to identify the molecules. The latest and most improved technique for Raman is surface-enhanced Raman spectroscopy (SERS). This technique can detect single molecules. This technique typically requires a silver or gold colloid or substrate. In most cases, samples are prepared as plasmosurfaces composed of silver/gold nanostructures on porous silicon wafers. Samples can be expensive and time consuming to prepare.

There are many advantages in using standard Raman spectroscopy. Raman can be used on solids and liquids, no sample preparation is required, water does not interfere with analysis, and Raman is non-destructive. This technique can locate chemicals very reliably and the analysis of this method is very rapid. Raman analysis is available for relatively small sample sizes (<1 μm) and inorganics are more readily detected by Raman than by IR spectroscopy. One of Raman's most significant advantages is that this type of analysis can be performed under standard atmospheric conditions.

Fourier transform infrared (FTIR) spectroscopy is a technique that utilizes the vibrational response of chemical bonds when compounds are exposed to the IR spectral range. Based on this technique, the infrared spectrum emitted or absorbed during exposure to the IR spectrum is observed for solids, liquids or gases and used to identify compounds. FTIR has advantages over Raman in that it is less of a problem with interference, eg fluorescence. However, FTIR inherently requires minimal thickness, uniformity and dilution.

Metrology techniques also analyze material properties of the substrate 4 and/or substances located adjacent to the substrate 4 before, during and/or after the removal process as shown in FIG. 13 or Used according to certain embodiments of the present invention to monitor. For example, measurements of the material properties of a partially absorbing film on a substrate can be used to calculate the phase retardation of the material prior to processing. This can determine the process temperature for cleaning to induce the appropriate phase retardation in the absorbent film. This metrology technique can also be used to monitor phases during processing and provide feedback information to the process or to stop the process if the feedback information is outside process limits. Material characterization results of the substrate can be used to calculate precise energies to induce desired surface material or morphological feature changes. Additionally, monitoring the material properties of the pellicle membrane can determine if adverse effects are occurring to the pellicle material. This can be used prior to processing to limit the process temperature, or the process can be stopped if damage is observed during processing. For example, one or more ellipsometers or cameras 31 can be used to measure the material response of the pellicle film, absorber film and substrate surface. This data can then be used to calculate desired material properties including film thickness, transmittance and phase.

Other metrology techniques for monitoring the presence and amount of surface contamination can be utilized prior to, during and/or after the cleaning process of the present invention. For example, metrology techniques used to detect the presence of contamination can be used in accordance with certain embodiments of the present invention to determine whether to apply a laser pulse to the measured area of the substrate. can do. This information can then be used to minimize the number of pulses applied across the substrate, thereby reducing the total thermal energy applied to the system as well as the overall cleaning process time.

Metrological methods for measuring the lateral size/dimension, location, number, density and/or height (thickness) of contaminants or contaminant particles are also used in certain implementations of the invention in combination with the cleaning process. Can be used according to form. For example, these measurements can be used to assess the efficiency and completion of the process by measuring before and/or after the cleaning process. During the process, these measurements can be used to assess the field efficiency of the process. For example, if a number of laser pulses were used to complete ablation, detection of residual contamination can be used to estimate the number of pulses required for ablation, if additional pulses are required. can. In this case, the metrology technology is configured to observe the area being cleaned while the cleaning process is taking place, according to certain embodiments of the invention. This is typically done by imaging the area exposed by the laser, which involves using the same optics used for laser delivery as shown in FIG. good. The imaging lens 32 allows detailed monitoring of the contaminant particles 3 via the partially reflective mirror 29, which allows simultaneous monitoring and cleaning.

A number of methods exist for detecting particles and estimating particle size in accordance with embodiments of the present invention. These methods include, for example, reflected and transmitted light intensity measurements, imaging or imaging, low-angle scattered light detection, interferometry, scanning electron beam, scanning tunneling microscopy, near-field microscopy, atomic force microscopy, and the like. mentioned. A number of methods can be combined according to certain embodiments of the invention to obtain additional information.

For photomasks, for example, a number of metrology techniques can be incorporated into the laser cleaning process according to one particular embodiment of the invention. For example, identifying the presence of certain contaminants (e.g., ammonium sulfate) on a photomask defines decomposition temperature requirements and, in some cases, allows the selection of laser energies high enough to perform the cleaning process. .

According to one particular embodiment of the invention, the transmitted light intensity is measured and the result is compared with a programmed structure for the absorbing film on the photomask surface. The difference between the programmed and detected features is then used to identify contamination. Additionally, aerial imaging measurements are used in accordance with certain embodiments of the present invention to evaluate the printing properties of photomasks. This method is typically used to assess the impact of contamination on photomask performance. This measurement can also be used in situ to assess damage to the absorbent layer caused by the cleaning process. This is particularly relevant for partially absorbing films where the thickness of the partially absorbing film has a direct relationship with photomask performance. According to certain embodiments of the present invention, a combination of scattered light detection and transmitted light detection detects irregular surface topography that differs from the typically smooth surfaces of photomasks and photomask films. This improves the discrimination of contamination.

Metrology techniques are also used according to certain embodiments of the invention to monitor properties of materials located adjacent to the surface being cleaned. For example, monitoring the temperature of the pellicle film above the photomask can reduce the risk of damage to the pellicle film. Permeability properties of the pellicle membrane can also be used to assess the impact of the cleaning process during or after the cleaning process. Additionally, detection of particles on the inside of the pellicle membrane may be performed prior to performing the cleaning process and/or may be used to detect loss of these particles during this process and/or the pellicle and/or Or it can be determined if there is a limit to the energy that is preferably used in the process to prevent potential substrate material damage.

As one of ordinary skill in the art practicing one or more embodiments of the present invention will appreciate, the examples of metrology techniques described above are intended to be all that are encompassed by the present invention. It doesn't mean there is. In contrast, these examples merely illustrate the use of metrology techniques within some methods of the present invention.

Contamination identification

As noted above, photomask substrates typically consist of a base substrate that is transparent to the radiation used in the lithographic exposure process and an optical thin film that partially or fully absorbs the exposure radiation. Since the optical properties of these substrates are important to their performance, it is important to keep their surfaces as free of chemical contamination as possible. It is also important that these surfaces be free of contamination due to the energy and wavelength of the exposing radiation. This radiation can cause degradation with residual surface contamination and, in some cases, the formation of gradual defects (haze) that adversely affect photomask performance.

Although a great deal of effort has gone into the photomask cleaning process to minimize surface contamination, there is always some level of residual contamination, which can vary from photomask to photomask. In addition to residual surface contamination, there may also be environmental contamination buildup. Variations in surface contamination of photomasks used in wafer fabrication can result in variations in photomask performance and lifetime during manufacturing. Therefore, it is desirable to be able to detect the presence and level of surface contamination and analyze its chemical composition.

Identification of contaminants removed from photomasks or other contaminated substrates can be useful in identifying and eliminating or reducing sources of contamination and in customizing removal methods to minimize the potential for substrate damage. . Contaminant identification involves liberating molecules from the surface of a contaminated substrate, taking into account different geometries and substrate materials, sample pretreatments, and chemical analysis of the contaminants.

A number of ways of performing such identification will now be described in connection with FIG. 17, where each method involves energizing the surface of the contaminated substrate 4, for example with a laser 31, to cause photons to strike the contaminated substrate surface. , as a result of which the molecule is either by the transfer of momentum from photon to molecule, by the cleavage of a photo- or thermal bond between the contaminant and the substrate surface, by a photon-induced chemical reaction in which the molecule is decomposed or made, or It is liberated either by pyrolysis or generation due to heat generated from the substrate. In one embodiment, when liberating contaminants, the temperature of the photomask is maintained below a threshold temperature to prevent damage to the photomask. In all methods, the molecules are brought into the gas phase near the substrate surface. A pressure differential may be created near the surface of the substrate 4 to force liberated molecules away from the substrate, which are then collected or captured for analysis. As will be appreciated by those skilled in the art, it is advantageous to create a pressure differential over a localized area of the substrate, particularly within the area where the laser is illuminating the substrate. In one embodiment, a cylindrical tube 38 is provided surrounding the biasing source 31 and extending to the surface of the substrate 4 . A sensor 39 located above the substrate can be used to provide input to the gas analysis tool 40 so that contaminants can be identified. Alternatively, a pressure differential may be created in the cylindrical tube 38, for example by a vacuum pump 41, and this pressure differential may be used to draw free-floating molecules into the analysis tool or to the collection device 43. is good. Detection or chemical analysis methods may vary depending on the molecules to be identified and whether the contaminants 3 are drawn directly into the analysis tool 47 or accumulated by the collection device 43 .

Liberation of molecules from the surface of a contaminated substrate can be accomplished using a number of methods, including thermal evaporation, pyrolysis, or physical momentum transfer, photolytic bond cleavage, or other methods to liberate surface contamination. of photolytic means. A preferred embodiment uses a laser. The laser wavelength may vary based on the substrate material and the contamination to be analyzed. At certain wavelengths, the material may be transparent, meaning that the photons do not have much effect on the substrate as they pass through it. At other wavelengths, the substrate and/or contaminants can act as a partial or black body absorbing some or all of the photons and energy from the energizing source. This results in elevated substrate and/or contaminant temperatures. At sufficiently high energies, photon densities and appropriate wavelengths, this can result in liberation of contaminants from the substrate surface. In preferred embodiments, the electromagnetic wave has a wavelength that is substantially the same as the local maximum of the absorption spectrum of the photomask.

The contaminant identification methods described herein identify airborne molecular contaminants (AMC) either when they are being liberated directly from the contaminated substrate or after they have accumulated within the collection device 43. You can analyze things. For AMC liberated into the atmosphere on contaminated substrates, the apparatus may employ any type of sensor 39 and gas analysis tool 40 . Alternatively, free AMC from contaminated substrates may be deposited in collection reservoir 44 or on collection substrate 45 . AMC liberated from contaminated substrates may be confined within container 44 or condensed onto the surface of collection substrate 45 in a concentrated state. This collection device 43 can collect AMC from a single contaminated substrate or many contaminated substrate samples before performing the analysis and if the collection substrate 45 can be analyzed using multiple surface analysis methods. can.

In one embodiment, the collection substrate 45 may be a cold plate that is kept below 0° C. to make condensation of vapor phase contaminants more likely. For example, the collection substrate can be cooled to −195° C. by liquid nitrogen N ₂ or cooled to the desired temperature by a Peltier cooling device. A pressure differential may be created to draw the AMC from the environment above the contaminated substrate 4 toward the cryogenic collection substrate, thereby increasing the likelihood of AMC condensation and deposition.

In another embodiment, collection vessel 44 may be a closed volume with inlet and outlet ports. The interior of this volume may be cooled below 0° C. to make condensation of vapor phase contaminants more likely. A pressure differential may be created to draw AMC from the environment above the contaminated substrate 4 through the inlet port and into the enclosed volume, thereby increasing the likelihood of AMC condensation and deposition.

It is very important to understand that any condensable gas in the incoming flow caused by the pressure differential will condense on the collection substrate 45 or within the collection vessel 44 when cooling the collection device. This includes any atmospheric water due to the humidity in the environment around the collector 43 . For this reason, it is recommended to place the contaminated substrate surface and collection device 43 in an environment where the humidity is reduced forward from this point, called a dry environment.

A dry environment can be achieved by replacing the air around the contaminated substrate and collection device with a dry gas, preferably a dry inert gas such as nitrogen, argon or helium. The dryness of the environment can be monitored by measuring the dew point, which should be kept below -10°C. This minimizes the amount of water from the atmosphere that condenses on the collection substrate or in the collection vessel, yet reduces the signal-to-noise ratio (signal-to-signal ratio) of metrology techniques performed on contaminants during identification and quantification. noise ratio) is improved.

Also, as will be appreciated by those skilled in the art, the proximity of the cooling collection device to the contaminated substrate surface is important. The closer the collector is to the contaminated substrate surface, the more AMC is deposited by the collector. This is important when dealing with small amounts of contaminants.

After depositing the AMC in the collection device, the AMC can then be analyzed using metrological techniques such as mass spectrometry. AMC can also be re-evaporated and analyzed in the gas phase using various gas phase spectroscopy techniques.

Methods used for analysis include, but are not limited to, spectroscopy, mass spectrometry, electrochemical analysis, thermal analysis, separation, hybrid techniques (including two or more techniques), and microscopy. not. For photomask substrates, some analytical techniques such as spectroscopy, mass spectrometry, separations and hybrid techniques have more advantages.

The design of contamination identification devices in accordance with aspects of the present invention may also vary depending on geometry. Some applications are better on flat substrate surfaces. This application may employ a stationary biasing source and a collection tube surrounding the biasing surface and the biasing source. Other applications may have geometrical obstructions, and the device may utilize adjustable contraction optics and telescoping collection tubes. If the contaminated substrate contains a foreign material, the apparatus may utilize an energizing source of varying wavelengths. Preferred embodiments allow for various geometries and multiple material substrates.

The use of the method of the present invention for analysis of surface contamination of photomasks depends on the point in the manufacture of the photomask. When the photomask is first manufactured, the entire substrate surface is exposed to the atmosphere. At this step in the fabrication of the photomask, a source of electromagnetic radiation can be applied over the photomask. The radiation can be exposed through the entire photomask at once or the radiation source can be focused toward the substrate surface and the surface and source can be scanned relative to each other. Free AMC contamination can be drawn from the atmosphere directly over the contaminated substrate and analyzed directly by metrological tools, such as by mass spectrometry. Alternatively, the AMC may be collected by a collection device and later analyzed.

Analyzing AMC as it is fabricated directly above the photomask substrate has advantages for certain embodiments of the present invention. For example, in one aspect of the invention, the presence, relative quantity and identity of contaminants is determined when the laser is first directed at the substrate. The contamination level can then be compared to the average level of the photomask collection or to a nominal acceptable value for contamination. The level of contamination and the type of contamination can be used to determine whether a reticle is clean enough to be used in manufacturing. The photomask can be removed from use or sent to another cleaning process to further reduce surface contamination. By identifying contaminants, the overall photomask manufacturing process can be adjusted to remove the contaminants, thereby improving overall manufacturing quality. The process of the present invention can be applied to the photomask for a second time to determine the level of contamination and the degree of discrimination. This measurement can be compared to previous levels and used in the process of the present invention to determine whether the cleaning process is reducing surface contamination. The cleaning process and contamination identification process can be used repeatedly until the desired level of surface contamination is reached.

This method has advantages compared to other direct metrology techniques. The process of the invention is not limited to a single analytical technique. Gas and substrate samples can be easily prepared, allowing the use of many powerful analytical tools. The method of liberating AMC is non-destructive and thus does not waste the product, eg photomask. The disclosed method does not require a vacuum chamber to perform the analysis. Most analytical techniques are limited by sample size. The process of the present invention is a local process for each energizing pulse bounded by the process area. However, due to the continuous pulses and movement of the contaminated sample substrate, the sample size is not limited, especially when incorporating high precision stages.

In another aspect of the invention, the electromagnetic radiation used to detect the presence, level and identity of AMC can assist in cleaning the surface of the photomask. In this case, surface contamination levels are expected to decrease when radiation is directed at the surface. Repeated exposures of several portions or the entire photomask surface can be used in combination with free contamination level monitoring to determine when the photomask is sufficiently clean to be used in manufacturing.

In the latter part of photomask fabrication, a pellicle is added onto the top surface of the photomask substrate. The pellicle consists of a hollow rectangular frame bonded to one face of the substrate and having a thin fluoropolymer membrane (pellicle) across the top. The presence of a pellicle makes the photomask incompatible with vacuum systems. For example, applying a vacuum to the photomask releases volatiles from the pellicle frame and glue, which further contaminate the photomask surface. The process of the present invention can be used on areas on the top surface of the photomask that are outside the pellicle frame or on the back side of the photomask without the pellicle. Alternatively, electromagnetic radiation may be used through a pellicle. However, only those species of AMC that are able to permeate the pellicle membrane may be detected and identified. The method of the present invention can be applied to the photomask immediately after it is manufactured or after the photomask has been in manufacturing use for some period of time. By measuring the level and identity of AMC immediately after the pellicle is utilized, surface contamination levels can be determined prior to use in manufacturing. Analysis of surface contamination levels after the photomask has been used in manufacturing can determine the level and type of environmental contamination to which the photomask was exposed during manufacturing use. This allows early detection and identification of surface contamination build-up before reticle damage occurs during manufacturing, while allowing the manufacturer to adjust processes to prevent further contamination.

Due to the typically low levels of contamination on photomasks, it may be advantageous to collect loose contamination from a single photomask. Depositing loose material from photomasks can improve the ability to determine the presence, level and identity of contamination. In addition, chemical analysis of surface contamination typically requires a larger sample volume than detection alone. Depending on the contamination level, collection of loose AMC from multiple photomask substrates onto a single collection substrate to deposit a large amount of residual surface contamination identifies the contamination and locates the source. can be very useful on

As will be appreciated by those skilled in the art, the contamination identification apparatus and methods disclosed herein allow for on-site or remote analysis of contamination on substrates. This analysis can be either destructive or non-destructive and can be used for chemical analysis, gas or surface analysis methods. Multiple analytical formats may be used simultaneously for more accurate discrimination or self-calibration. Also, as will be appreciated by those skilled in the art, the collected sample should have a much higher concentration than the contaminated substrate surface, allowing detection of low levels of contaminants. A chemical analysis can be performed on a large number of contaminated substrate surfaces and monitored over time, the chemical analysis being performed on a large number of contaminated substrates with graded materials and geometries. can be done.

Device

Certain methods of embodiments of the present invention are incorporated into equipment used to perform laser surface cleaning processes. An example of such an apparatus is shown in FIG. 15, which embodiment includes a robot 35 and one or more motion-struggling robots 35 that handle and precisely position the substrate material using robotic end effectors. It also has a platform 34 for positioning the substrate sample relative to the laser beam. The apparatus may include, for example, one or more of the metrology techniques described above and/or include ways to control the temperature of the substrate and/or adjacent materials during the cleaning process. is good. Additionally, the apparatus may include metrology techniques used to position the substrate relative to the staging system and thus the laser beam. The metrological technique may further include a computer controlled visual recognition system. Additionally, the apparatus may utilize computerized control of laser, motion and/or metrology techniques and may provide software-based recipe control of the cleaning process. Laser control includes, for example, controlling when laser pulses are applied as well as the amount of energy applied during the cleaning process.

Wafer fabrication process

A method and/or apparatus according to certain embodiments of the present invention can be utilized as part of a novel wafer fabrication process involving the removal of haze formations from a pellicled photomask surface. Typically, the photomask is removed from the wafer print process when the level of haze is sufficient to adversely affect the wafer print process. The time before removing the photomask is typically determined either by direct detection of high levels of haze contamination or based on predetermined durations and/or usage levels during wafer processing. be. Typically, the photomask is sent to another facility to remove the pellicle, clean the photomask, and attach another pellicle to the photomask. These other facilities (eg, mask factories) maintain the equipment necessary to accomplish these tasks and perform additional inspections not required at photomask repair and wafer fabrication facilities. Typically, two identical sets of photomasks are used for the time required to clean the photomasks and install a new pellicle. These additional photomasks add significantly to the cost of the overall wafer printing process by requiring high material and set-up and evaluation costs.

A novel wafer fabrication method as a particular embodiment of the present invention incorporates apparatus that utilizes one or more of the methods described above to remove haze from a photomask surface. A representative wafer fabrication process as an embodiment of the present invention illustrates the use of a wet cleaning process to remove photomask contamination as shown in FIG. 16A. Another method also included within the scope of certain embodiments of the present invention utilizes one or more of the laser cleaning methods described above to perform cleaning operations in a wafer fabrication facility, Removal is not shown in the flow diagram of FIG. 16B. This can minimize or eliminate additional pellicle cost and/or photomask film degradation resulting from current wet cleaning processes.

According to certain embodiments of the present invention, the novel wafer fabrication process eliminates the use of additional masks or mask sets for product fabrication while the original set is being cleaned. In this manufacturing process, the original photomask is immediately returned to production following the cleaning process shown in the flow chart of Figure 16C. This has the potential to eliminate the need for two identical masks and reduce the required set-up time for using two identical mask sets. The use of inspection metrology to verify the cleaning process can advantageously be done before the photomask is returned to production. This measurement may, for example, be built into the equipment or provided by additional equipment at the wafer fab or other facility. Overall process time for photomask haze removal is reduced despite metrology techniques.

Many features and advantages of the present invention will be apparent from the detailed description, and it is thus intended to cover all such features and advantages of the invention that fall within the true spirit and scope of the invention to the appended claims. is intended by the description of the range of Furthermore, since many modifications and variations will readily occur to those skilled in the art, it is not desired to limit the invention to the exact construction and operation shown and described; Suitable modifications and equivalents of are available.

1 laser 2 excitation energy 3 contaminant particles 4 substrate 5 decomposed particles 6 vaporized material 7 thin film absorber 8 pellicle 9 pellicle frame 10 adhesive 11 focusing lens 12 converging lens 13 laser pulse or mask beam 14 pellicle beam 17 spatial distribution 18,19 Heat Exchange Tube 21 Side Flow 20 Bottom Flow 23 Pattern

Claims (37)

A method of identifying contaminants on a surface of a photomask, comprising:
directing electromagnetic radiation from an electromagnetic radiation source toward a photomask having contaminant particles attached thereto, said electromagnetic radiation having a wavelength substantially identical to a local maximum of an absorption spectrum of said photomask;
causing a temperature rise in the photomask;
transferring thermal energy from the photomask to the contaminant to liberate free molecules of the contaminant from the surface of the photomask;
creating a pressure differential over the photomask to force loose molecules of the contaminant away from the photomask;
trapping free molecules of said contaminants on a collection substrate cooled to −195° C. with liquid nitrogen _N2 ;
A method comprising analyzing the composition of said contaminant. 2. The method of claim 1, wherein the electromagnetic wave is laser light. 3. The method of claim 2, wherein the wavelength of said laser light is greater than 8 micrometers. 2. The method of claim 1, wherein the photomask has at least one thin film layer. 5. The method of claim 4, wherein said thin film layer is patterned, said thin film layer having void regions, and wherein respective portions of said photomask are exposed under said void regions. 2. The method of claim 1, further comprising maintaining the temperature of the photomask below a threshold temperature to prevent damage to the photomask. 3. The method of claim 2, wherein the photomask comprises at least two materials and the wavelength of the laser light is substantially the same as the local maximum of the absorption spectrum of the substrate. 2. The method of claim 1, wherein the pressure differential is localized over less than the entire photomask. 3. The method of claim 2 , wherein the pressure difference is localized to the area where the laser light impinges on the photomask. 2. The method of claim 1, wherein said pressure differential is created within a collection tube around said source of electromagnetic radiation to said photomask. 3. The step of capturing free contaminant molecules comprises cooling a collection substrate to below 0° C. and using the pressure differential to direct the free contaminant molecules to the collection substrate. 1. The method of claim 1. 12. The method of Claim 11, wherein the photomask and the collection substrate are located in a dry environment at atmospheric pressure. 2. The method of claim 1, wherein analyzing the composition of the contaminant utilizes spectroscopy. 2. The method of claim 1, wherein analyzing the composition of the contaminants utilizes mass spectrometry. 2. The method of claim 1, wherein analyzing the composition of the contaminants determines the level and type of contaminants. 2. The method of claim 1, wherein analyzing the composition of the contaminants further comprises comparing the level and type of contaminants to a benchmark for the photomask. A method of identifying contaminants on a surface of a substrate, comprising:
directing an electromagnetic wave toward a substrate having contaminant particles attached thereto located in a dry environment at atmospheric pressure, wherein the electromagnetic wave is substantially equal to a local maximum of an absorption spectrum of the substrate; have the same wavelength,
causing a temperature rise in the substrate;
transferring thermal energy from the substrate to the contaminant to liberate free molecules of the contaminant from the surface of the substrate;
trapping free molecules of said contaminants on a collection substrate cooled to −195° C. by liquid nitrogen _N2 and placed in a cooled collector;
A method comprising analyzing the composition of said contaminant. 18. The method of claim 17, further comprising maintaining the temperature of the substrate below a threshold temperature to prevent damage to the substrate. 18. The method of claim 17, wherein the cooled collection device is a cold plate. 18. The method of claim 17, wherein the refrigerated collector is cooled below 0<0>C. 18. The method of claim 17, wherein the cooled collection device is positioned in close proximity to the substrate surface. 18. The method of claim 17, wherein the refrigerated collection device is a refrigerated enclosed volume with an inlet port and an outlet port. A pressure differential is created between the cooled enclosed volume and the dry environment such that free molecules of the contaminant enter the cooled enclosed volume through the inlet port. 23. The method of claim 22. 18. The method of claim 17, wherein the dry environment has a dew point of less than -10°C. 18. The method of claim 17, wherein the dry environment is an inert gas environment. 26. The method of claim 25, wherein said inert gas is nitrogen. A method of cleaning a surface of a substrate, comprising:
directing an electromagnetic wave from an electromagnetic wave source toward a substrate having a contaminant attached thereto located in a dry environment at atmospheric pressure, the electromagnetic wave being at a local maximum of an absorption spectrum of the substrate; having substantially the same wavelength,
causing a temperature rise in the substrate;
transferring thermal energy from the substrate to the contaminant to decompose the contaminant;
trapping the decomposed contaminant molecules on a collection substrate cooled to −195° C. by liquid nitrogen _N2 ;
A method comprising analyzing the composition of said contaminant. 28. The method of claim 27, wherein the electromagnetic wave is laser light. 29. The method of Claim 28, wherein the wavelength of said laser light is greater than 8 microns. 28. The method of claim 27, wherein the substrate has at least one thin film layer. 31. The method of claim 30, wherein the thin film layer is patterned, the thin film layer having void regions, and wherein respective portions of the substrate are exposed under the void regions. 28. The method of claim 27, further comprising maintaining the temperature of the substrate below a threshold temperature to prevent damage to the substrate. 29. The method of Claim 28, wherein the substrate comprises at least two materials and the wavelength of the laser light is substantially the same as the local maximum of the absorption spectrum of the substrate. 28. The method of claim 27, wherein the dry environment has a dew point of less than -10°C. 28. The method of claim 27, wherein the dry environment is an inert gas environment. 36. The method of claim 35, wherein said inert gas is nitrogen. 28. The method of claim 27, wherein the substrate is placed in a dry environment at atmospheric pressure before surface cleaning begins.

Images