Optical particle counter and method

Abstract

This specification relates to an optical particle counter, a method for using an optical particle counter, and a method for reducing particle counting errors during use of the optical particle counter. The method includes a filter placed between a laser source and the fluid to be measured. The filter may be positioned at an angle not normal to the laser.

Description

Technical Field

[0001] This specification relates to an optical particle counter, a method of using the optical particle counter, and a method of reducing particle counting errors during use of the optical particle counter.

[0002] Cross - Reference to Related Applications This application claims the benefit under 35 U.S.C. § 119(e) of U.S. Provisional Patent Application No. 63 / 471,405, filed on Jun. 6, 2023, the disclosure of which is incorporated herein by reference in its entirety.

Background Art

[0003] Optical particle counters have various commercial and industrial uses for identifying the presence or amount of specific types of particles in a fluid. The fluid may be a liquid (e.g., aqueous or organic liquid) or a gas (e.g., clean air in a “clean room” or exhaust gas).

[0004] In certain applications, the fluid is a “slurry” (also called a “particle dispersion” or “particle suspension”) that contains small particles dispersed or suspended in a liquid medium. Specifically, the slurry may be of the type used in a CMP process (“chemical - mechanical planarization” or “chemical - mechanical polishing”) to treat the surface of a semiconductor wafer or micro - electronic device (“CMP substrate”). In these applications, special optical particle counters are used to detect relatively large particles (e.g., agglomerated particles) present in the slurry.

[0005] CMP slurries contain small abrasive particles dispersed in an aqueous medium. The slurry is used to planarize the surface of a semiconductor wafer by applying the slurry to the surface and using a moving pad (“CMP pad”) to contact the pad and slurry against the surface with pressure. Pressing the slurry and pad against the surface abrasively wears away a small amount of material from the semiconductor wafer surface, thereby producing a very flat, i.e., “planarized” surface.

[0006] The particle size in the slurry can be selected to provide a desired polishing effect. The slurry can be prepared to have particles within a desired size range, often with a particle size distribution that has the morphology of a typical (i.e., bell-shaped or Gaussian) curve. However, over time, after the slurry is prepared and before it is used, the individual particles in the slurry tend to coalesce, forming larger "aggregated" particles. These aggregated particles can be large enough to generate scratches on the semiconductor wafer surface during the planarization step.

[0007] An optical particle counter can be used to detect undesirable particles in a slurry, including aggregates formed from multiple individual polishing particles. When used to detect aggregates in a slurry, the optical particle counter directs focused laser light through the slurry sample as the sample moves through a small transparent conduit called a "flow cell." The laser light passing through the moving slurry may be reflected or interrupted as aggregates pass through the light. The particle counter includes one or more photodetectors positioned relative to the moving slurry sample, which receive the light from the laser passing through or reflected by the slurry. Changes in the intensity of the laser light received by the detector may indicate undesirable large particles in the slurry, such as aggregates. [Overview of the project]

[0008] Optical particle counters are highly sensitive measuring devices that rely on a precise combination of highly advanced optical components such as lasers, transmissive optical devices such as flow cells, optional filters, electronic components such as photodetectors and electronic control systems, and mechanical components. The combined components are controlled and monitored by computer software, which must be programmed and applied to the specific measurement application to accurately interpret the electronic signals generated by the photodetector, which receives and measures the intensity of laser light passing through or reflected by the slurry. The system can be highly sensitive to very small differences in operating conditions, variability in the performance of system components (e.g., consistency of laser wavelength and output, which can be affected by laser temperature and current input), and the properties of the slurry and its components (liquid medium and particles). One or more changes or variations in these can cause erroneous readings, i.e., "particle counting errors," in the optical particle counting system.

[0009] When an optical particle counter produces a particle counting error, the cause of the error may be obvious or it may be latent and extremely difficult to identify from among the numerous variations, inputs, and operating conditions that combine to produce the measurement result. When an optical particle counter produces a reading that is suspected to be erroneous, there may be many potential causes of the error. Identifying the specific cause of the error can be an extremely difficult task requiring a thorough examination of all the different optical, computer, and electronic components of the instrument.

[0010] As described herein, certain types of particle counting errors can be caused by laser "mode hopping." In addition, in addition to many other potential causes, mode hopping can be caused by "optical feedback," i.e., light from a laser used to perform particle size measurements of a slurry is scattered and returned to the laser from the particles in the slurry. When certain types of optical particle counters are used to measure the presence of undesirably large particles in a slurry, such as a CMP slurry, the light guided into the slurry may hit small particles (micron or nanoscale particles) in the slurry, be scattered, and returned in large quantities to the laser. The scattered and re-induced light re-incident to the laser can induce an effect known as "mode hopping" of the light emitted by the laser. Laser mode hopping has the effect of causing variability in the light emitted from the laser, such as undesirable fluctuations in the intensity or frequency of the emitted laser light. When fluctuating (irregular) laser light is generated by the laser of an optical particle detector, the optical sensor of the optical particle counter receiving that light from the laser may produce erroneous signals, such as signals indicating a much higher concentration of relatively large particles detected in the slurry. Potentially, the error could cause the optical particle counter to report results (particle counts) that are more than 100 times higher than the expected or known particle count of the slurry.

[0011] In one embodiment, the present invention relates to an optical particle counter. The counter comprises a flow cell adapted to accommodate a flow of liquid dispersion; a laser that generates a beam of laser light directed towards the flow cell; an attenuation filter located between the laser and the flow cell, adapted to reduce the intensity of the laser light as it passes from the laser to the flow cell and also to reduce the intensity of the laser light returned from the flow cell to the laser; and an optical detection system for detecting the laser light leaving the flow cell.

[0012] In another aspect, the present invention relates to a method for detecting particles in a liquid dispersion using an optical particle counter. The method includes providing a flow of a liquid dispersion through a flow cell, wherein the liquid dispersion comprises a liquid medium having particles dispersed in a liquid medium; using a laser to generate a beam of laser light; passing the laser light through a flow of liquid dispersion in the flow cell; passing the laser light through a filter before passing it through the flow of liquid dispersion, the filter reducing the intensity of the laser light reaching the flow cell and also reducing the intensity of the laser light returning to the laser, thereby potentially generating optical feedback that could otherwise cause mode hopping by the laser; and detecting the laser light leaving the flow cell.

[0013] In yet another aspect, the present invention relates to a method for reducing laser mode hopping in an optical particle counter comprising a flow cell comprising a liquid dispersion comprising particles dispersed in a liquid, wherein the laser guides laser light into the flow cell and the liquid dispersion, a photodetector detects electromagnetic radiation leaving the flow cell, and the method comprises placing an attenuation filter between the laser and the flow cell to reduce the intensity of the laser light entering the flow cell and to reduce the intensity of the light returned to the laser as optical feedback which may cause mode hopping. [Brief explanation of the drawing]

[0014] [Figure 1] This is a diagram of a conventional optical particle counter. [Figure 2] This is a diagram showing the optical particle counter described. [Figure 3] This figure shows particle counting data generated from a conventional optical particle counter, as shown in Figure 1. [Figure 4] This figure shows particle counting data generated from a conventional optical particle counter, as shown in Figure 1. [Figure 5] This figure shows particle counting measurement data, including particle counting errors. [Figure 6]This is a decision tree diagram showing possible causes of particle counting errors. [Modes for carrying out the invention]

[0015] The following describes optical particle counters that are useful for detecting and measuring the presence of specific types of particles in liquid slurries.

[0016] An optical particle counter (or “counter”) may be useful for detecting relatively large particles (e.g., relatively large “aggregates”) in a slurry containing concentrations of these larger particles along with smaller particles suspended in a liquid medium. The slurry may contain any type of liquid medium and any type of solid particles. The following description relates to one specific example of the application of the optical particle counter and method of the present invention for detecting certain relatively large particles that may be present in a slurry used for chemimechanical planarization of semiconductor substrates.

[0017] Chemical mechanical planarization (CMP) is a process that uses chemical and mechanical forces together to polish a surface. It is used by semiconductor device manufacturers in various specific processes to prepare semiconductor wafers for the manufacturing process. Before the deposition or fabrication of complex micro and nanoscale semiconductor components, the working surface of a substrate (e.g., a 300mm semiconductor wafer containing microelectronic devices) must meet tight tolerances for flatness and smoothness to maximize quality and repeatability. The process ("planarization") is carried out using a CMP slurry, which is a liquid dispersion containing a certain concentration of small (e.g., micron or nanoscale) abrasive particles suspended ("dispersed") in a liquid medium.

[0018] Abrasive particles can be any type of abrasive particle useful in chemical machining compositions. Examples include various forms of zirconia, alumina, ceria, silica, and other ceramic materials. None of these may be doped or undoped, and they may be prepared by any of the various known methods for different types of particles. The abrasive particles are dispersed or suspended throughout a liquid medium. Various types of abrasive particles (e.g., charged, aggregated, uncharged, non-aggregated) are known and commercially available. Individual particles may be spherical or nearly spherical in shape, but they can also have other shapes such as elliptical, square, or rectangular cross-sections. Individual particles are called primary particles. Generally, individual (primary) particles cluster or combine to form larger particles ("aggregated particles") from multiple individual particles.

[0019] Aggregated particles, consisting of several to several individual particles (e.g., 2, 3, 5, or 10 particles), are common and useful in slurries. However, aggregated particles can grow to a size large enough to cause damage to the substrate during processing. In CMP slurries, abrasive particles can coalesce to form larger aggregated particles due to pH shifts, shear stress, or temperature effects. Some particles may become excessively large, taking the form of large "aggregated" or "gel" particles. In slurries, the size of individual abrasive particles and acceptable aggregated particles can be submicron, e.g., on the nanometer scale. Useful abrasive particles (individual or acceptable aggregated particles) may have an average particle size ranging from about 0.1, 0.5, 1, 5, 10, or 20 microns to about 100 microns. Larger, undesirable aggregated particles are larger than the sizes of useful abrasive particles, e.g., larger than D90, D95, or D99 of useful particles, and may be large enough to damage the substrate during processing.

[0020] The slurry analyzed by the described system may contain any appropriate amount of abrasive particles, e.g., about 0.01, 0.05, 1, 2, or 5% by weight to about 10, 20, or 30% by weight based on the total weight of the slurry. The particles have a particle size distribution that includes a high proportion of all particles having particle sizes in the micron range, for example, at least 90 or 95% of the particles in the slurry have a particle size of less than 100, 50, 10, 1, 0.5, or 0.1 microns, for example, at least 90 or 95% of the particles have a particle size of less than 5, 1, 0.5, or 0.1 microns, i.e., measured D90 or D95 of less than 100 microns, or measured D90, D95, or D99 of less than 5, 1, 0.5, or 0.1 microns.

[0021] The slurry typically comprises abrasive particles dispersed in a liquid medium containing a high concentration of water and any small amount of organic additive such as a solvent, the solvent of which may, among other things, be a lower alcohol (e.g., methanol, ethanol, etc.) or an ether (e.g., dioxane, tetrahydrofuran, etc.). An exemplary liquid carrier may contain, or essentially consist of, water, more preferably deionized water. A carrier consisting essentially of water may contain up to (or less) 3, 2, 1, 0.5, 0.1, or 0.05 wt% of another liquid, such as a non-aqueous (organic) solvent such as a lower alcohol (e.g., methanol, ethanol, etc.) or an ether (e.g., dioxane, tetrahydrofuran, etc.).

[0022] To achieve the high-quality requirements and high process yield requirements necessary for current microelectronic device manufacturing, all variables present in the manufacturing steps, including the composition of the CMP slurry used during the planarization step, must be monitored and controlled as tightly as possible. Potential variations in the CMP slurry include the concentration and distribution of polishing particles within the slurry, particularly the presence and concentration of agglomerated particles of a larger size. Large agglomerated particles in the CMP slurry (e.g., agglomerated particles having a size greater than the D90, D95, or D99 of the useful slurry particles, e.g., greater than 0.5 microns) can cause scratches and dig defects on the surface of the wafer during CMP processing. As a result, it is important to monitor the size distribution and concentration of particulate matter in the CMP slurry, particularly the presence of agglomerated particles of an undesirably large size.

[0023] Generally, an optical particle counter uses a focused laser beam to detect particles contained in a liquid slurry. The laser beam is directed at a moving sample of the slurry and passes through at least a portion of the slurry. The light of the laser beam may pass through the entire slurry or may be blocked, scattered, or reflected by specific particles in the slurry as the sample passes through the laser beam. Certain small particles, such as useful micron and nanoscale particles in the slurry, may not block or reflect the light of the laser, but larger-sized particles, e.g., larger-sized aggregates having a particle size greater than the useful slurry particles, may block the light or may reflect, deflect, or scatter the light.

[0024] Photodetectors arranged around the sample collect the light that passes through the slurry or that is reflected, scattered, or deflected by the particles of the slurry. The particle counter uses information from the transmitted light or from the scattered or deflected light to calculate the amount or size of the particles in the slurry sample.

[0025] An exemplary system includes a photodetector on the opposite side of the laser that detects light from the laser after the beam has passed through a slurry. In these exemplary systems, particles in the slurry that are large enough to pass through the path of the laser beam will block the laser light, causing a temporary decrease in the light intensity. The particles are recognized by the temporary decrease in the intensity of the laser light detected by the photodetector. This type of system is called a “quenching” optical particle counting system.

[0026] Another type of exemplary system includes one or more photodetectors positioned relative to a slurry sample, enabling the detectors to receive light from a laser that is scattered or deflected by relatively large particles in the slurry sample. In these exemplary systems, relatively large particles in the slurry passing through the path of the laser beam are recognized as scattered or deflected light that scatters or deflects the laser light and is detected by photodetectors configured to sense this type of scattered or deflected light. This type of system is called a “scattering” particle detection system.

[0027] Referring to Figure 1, a conventional optical particle counter 2 is shown for detecting relatively large (e.g., very large aggregates) particles in a CMP slurry containing undesirable large aggregate particles in combination with smaller, useful, micron- or nanoscale slurry particles. The flow cell 30 is made of an optically transparent material and includes vertical sidewalls that form the sides of a channel adapted to accommodate a flow of sample CMP slurry 35. The flow cell may be made of a transparent material such as quartz or sapphire. An exemplary flow cell may define an internal channel having a rectangular cross-section with width and thickness dimensions useful for allowing the flow of CMP slurry through the channel during the process of counting particles using the optical particle counter. The exemplary thickness of the internal flow channel (in the direction aligned with the laser as it enters the channel) may be in the range of 200 to 400 microns (um). The exemplary width (in the direction across the laser as it enters the channel) may be in the range of 1.5 to 2.0 millimeters (mm).

[0028] The laser 10 generates a laser light beam 15 directed to pass through the lens 18 and the flow cell 30. The laser light passes through the slurry and useful micron or nanoscale particles of the slurry, but is interrupted by relatively large aggregated particles. The light that passes through the flow cell 30 without interruption is detected by the photodetector 20. In response to the light received by the photodetector 20, the detector 20 generates an electronic digital or analog output signal correlated with the intensity of the received light.

[0029] During use, relatively large aggregated particles in the CMP slurry passing through the flow cell 30 partially block the laser beam 15 as the beam passes through the flow cell 30 and the sample 35. This interruption temporarily reduces the intensity of the laser beam 15 received by the photodetector 20. By monitoring the intensity of the light from the laser beam received by the photodetector 20, the operating software of the optical particle counter 2 can detect, monitor, and report the presence and concentration of relatively large aggregated particles present in the sample CMP slurry 35 as the sample passes through the flow cell 30.

[0030] The illustrated optical particle counter 2 includes a post-cell optical filter 40 for attenuating the optical power of the laser beam 15. The post-cell optical filter 40 is used to attenuate (i.e. reduce) the power of the laser beam 15 received by the photodetector 20, based on the optical and electronic capabilities of the photodetector 20. For example, the photodetector 20 can exhibit a sensitivity range and effective identification at optical power levels lower than those produced by the laser 10. Therefore, the post-cell optical filter 40 is useful for reducing the optical power of the laser beam 15 received by the photodetector 20 to a level suitable for the capabilities of the photodetector 20. As shown in the illustration, the filter 40 is a type of attenuation filter called a “neutral density” filter or “ND” filter, which similarly reduces or modifies the intensity of light at all wavelengths.

[0031] The laser 10 and detector 20 can be lasers and detectors useful for optical particle counters, respectively, as shown in the figure. Examples of useful lasers may include single-mode lasers that emit single-frequency light with wavelengths in the range of 600–700 nanometers. Examples are commercially available, such as semiconductor laser diodes that emit light at a wavelength of 635 nanometers with power outputs in the range of 7–15 milliwatts. Examples of useful photodetectors 20, such as the Hamamatsu S2506-02 and Hamamatsu S6967, are also commercially available.

[0032] An optical particle counter, such as the one shown in Figure 1, can be highly sensitive to even very small differences in operating conditions, variability in the performance of system components (e.g., consistency of laser wavelength and output, which can be affected by laser temperature and current input), and the properties of the slurry and slurry components (liquid medium and particles). A change or variation in one or more of these characteristics of counter 2, as well as other conditions and characteristics of its operation, can cause counter 2 to produce an incorrect reading, i.e., a "particle counting error."

[0033] For example, Figures 3 and 4 are charts showing examples of data output from optical particle counters, such as counter 2 in Figure 1. The x-axis represents elapsed time, and the y-axis represents backlight intensity. Figure 3 shows a single peak indicating the presence of a single relatively large (oversized) aggregated particle detected in the slurry sample. The Gaussian-shaped peak is expected from actual particles passing through the laser beam. However, Figure 4 shows a square-wave-like peak that is not generated by actual particles.

[0034] If counter 2 operates in a way that produces particle counting errors, as shown in Figure 4, for example, the system is not effective for the purpose of detecting large aggregated particles. Particle counting errors are over-reporting of relatively large aggregated particles in the slurry sample. This false over-reporting of excessively large particles in the slurry that are not actually present necessitates the unwarranted exclusion of the monitored CMP slurry.

[0035] Generally, the causes of particle counting errors are either obvious or latent, and it is extremely difficult to identify them among the numerous variations in inputs and operating conditions of the optical particle counter that are combined to produce the measurement results. If an optical particle counter produces readings that are suspected of being erroneous or are known to be erroneous (for example, by being a measure of a known sample), there can be numerous potential causes of the error. Identifying the exact cause of a particle counting error can be an extremely difficult task, requiring a close examination of many different optical components, computer and software programming, as well as all the electronic components of the optical particle counter, including operating conditions (temperature) and all operating inputs (e.g., the current used to operate the laser).

[0036] When attempting to identify the cause of particle counting errors, the operation of the various optical components of the counter, individually or in combination, can be one of many potential causes. Laser operation is one possible cause, including how conditions (temperature) and input (current) can affect the optical output. Lasers are highly sensitive to variations in input and operating conditions, including operating temperature ("laser case temperature"), variations in input current, and "optical feedback," which is light entering the laser from outside the laser.

[0037] Optical feedback can affect the performance of a laser by altering the properties of the light emitted from the laser. Under certain conditions, the light emitted by and returned to the laser can cause the laser to experience "mode hopping," which means that the laser's optical output shifts from one longitudinal mode (wavelength) to a different mode. Mode hopping can result in irregular changes in the wavelength of the light emitted from the laser, along with intensity fluctuations (intensity "noise"). There are various potential causes of mode hopping, including the operating conditions of the laser (e.g., the laser's case temperature and current), the variability of the current used to operate the laser, and optical feedback.

[0038] The applicant determined that when using an optical particle counter as shown in Figure 1 to detect large aggregated particles in a slurry (e.g., CMP slurry) containing mostly smaller micron and nanoscale particles, the light emitted from the laser and directed towards the flow cell and slurry can be scattered by the smaller particles in the slurry and return to the laser in sufficient quantities to cause mode hopping, which can generate particle counting errors such as multiple false signal peaks identified in Figure 4.

[0039] Mode hopping is a well-understood phenomenon, known to be caused by laser light being returned to the laser; however, a counter 2 as shown in Figure 1 is not necessarily, or immediately, considered to require protection from optical feedback or mode hopping caused by light being returned from the slurry 35 to the laser 10. Mode hopping due to optical feedback is not an obvious cause of particle counting error in the system of Figure 1. Other possible causes of particle counting error in the form of overcounting include the focus shape of the laser beam, the laser temperature which may cause mode hopping, or the beam not being centered in the flow cell. Thus, identifying the apparent particle counting error, it is expected that many other operating conditions and inputs of the counter 2 are likely to be causes of the error, as opposed to mode hopping caused by optical feedback resulting from scattering and redirection of the laser light by micron or nanoscale particles in the slurry 35. See Figure 6 and the relevant description herein.

[0040] The applicant also determined that the particle counting error shown in Figure 4, caused by mode hopping, can be prevented by using an attenuation filter placed between the laser 10 and the flow cell 30 of the counter 2 shown in Figure 1.

[0041] More generally, when using an optical particle counter (e.g., the type shown in Figure 1) as specified by the applicant and described herein, to detect relatively large undesirable aggregated particles in a slurry containing mostly micron or nanoscale particles, optical feedback can be generated by the scattering of light emitted from a laser by the slurry and its return to the laser as optical feedback. Optical feedback can generate mode hopping, which can be the cause of particle counting errors generated by the optical particle counter.

[0042] According to the optical particle counter described herein, the counter may include an attenuation filter, or "precell filter," placed between the laser and the flow cell. The "precell" attenuation filter reduces the intensity of the laser light passing through the filter in two directions. First, the filter attenuates the light passing from the laser toward the flow cell, thereby reducing the intensity of the laser light that passes through the flow cell and is ultimately received by the optical sensor. The attenuation filter placed between the laser and the flow cell also attenuates the light emitted by the laser, scattered and re-induced by the slurry, otherwise it would become optical feedback by entering the laser. The reduced optical feedback prevents mode hopping.

[0043] The filter may be any useful attenuation filter, such as a “neutral density” filter or “ND” filter, which reduces or modifies the intensity of all wavelengths of light equally. Other examples are filters sometimes called “laser line filters,” “edge filters,” or “rejection filters.” An exemplary precell filter may be effective in reducing the intensity of light by at least 50, 80, 90, or 95% as light passes through the filter in one direction. A precell attenuation filter reduces the intensity of light passing through the filter for the first time in a first direction from the laser to the flow cell by at least 50, 80, 90, or 95%. The same precell filter also reduces the intensity of light passing through the filter a second time in a second direction from the flow cell to the laser, which is reflected by the slurry, by at least 50, 80, 90, or 95%.

[0044] Figure 2 shows two possible orientations of the filter 50 with respect to the passing laser beam 15. In the first position, shown by the solid line, the filter 50 is perpendicular to the laser beam 15. In the alternative position shown by the dashed line, the filter 50 may be slightly inclined with respect to the direction of the beam 15, for example, at an angle of 1 to 30 degrees away from perpendicular to the beam 15. An angled filter 50 can reduce reflection from the filter surface to the laser 10 and further reduce the possibility of optical feedback.

[0045] Referring to Figure 2, an optical particle counter 52 is shown for detecting very large agglomerated particles in a CMP slurry containing agglomerated particles in combination with smaller useful micron or nanoscale particles. The counter 52 can be adapted to detect agglomerated particles in a CMP slurry that are larger than 80, 90, 95, or 99% of the useful slurry particles, for example, to detect agglomerated particles having a size larger than D80, D90, D95, or D99 of the useful slurry particles. For slurries containing at least 90, 95, or 99% of particles smaller than 0.7 microns, the counter can be adapted to detect agglomerated particles larger than 0.7 microns, for example. For slurries containing smaller useful particles, the counter may be adapted to detect particles larger than 0.5 microns, or larger than 0.2, 0.15, 0.1, 0.5, 0.7, or 0.8 microns, up to approximately 20 microns, without detecting smaller useful slurry particles, such as those smaller than 0.8, 0.7, 0.5, 0.1, 0.15, or 0.2 microns.

[0046] Referring to Figure 2, particle counter 4 includes a similar arrangement of components to counter 2 in Figure 1, except for the location of an attenuation filter 50, which is shown as an attenuation filter 40, located between the laser 10 and the flow cell 30. As with counter 2 in Figure 1, the laser 10 generates a laser light beam 15 that passes through the flow cell 30. The laser light passes through the slurry and useful slurry particles of the micron or nanoscale, but is temporarily blocked by larger aggregated particles. The light that passes through the flow cell 30 without interruption is detected by a photodetector 20, which senses interruptions in the laser beam 15 caused by larger aggregated particles.

[0047] During use, large or aggregated particles in the CMP slurry passing through the flow cell 30 partially block the laser beam 15 as the beam passes through the flow cell 30 and the sample 35. This interruption temporarily reduces the intensity of the laser beam 15 received by the photodetector 20. By monitoring the intensity of the light from the laser beam received by the photodetector 20, the operating software of the optical particle counter 2 can detect, monitor, and report the presence and concentration of large aggregated particles present in the sample CMP slurry 35 as the sample passes through the flow cell 30.

[0048] The illustrated optical particle counter 4 includes a precell attenuation filter 50 positioned between the laser 10 and the flow cell 30 to reduce the intensity of the laser light beam 15 before it enters the flow cell. The precell filter 50 has the effect of attenuating (i.e. reducing) the output of the laser light beam 15 that enters the cell 30 from the laser 10 through the slurry 35 and is ultimately received by the photodetector 20. The precell filter 50 also has the effect of attenuating (i.e. reducing) the light (17) that is re-induced by the slurry 35 returning from the cell 30, which would otherwise enter the laser 10 as feedback. [Examples]

[0049] Table 1 shows a comparison of the results of detecting relatively large aggregated particles (>0.8 microns) in a slurry using particle counters arranged as shown in Figures 1 and 2. The slurry contains unwanted aggregated particles in an aqueous liquid medium from the concentrate, as well as useful smaller (micron or nanoscale) abrasive particles, and is mixed with either deionized water or filtered distilled water. The post-cell filter apparatus in Figure 1 includes an attenuation filter placed between the flow cell and the photodetector. The pre-cell filter apparatus in Figure 2 includes an attenuation filter placed between the laser and the flow cell. TIFF2026522557000002.tif51170

[0050] The particle count generated by the counter in Figure 1 was determined to include particle counting error, based on separate particle count measurements of the slurry.

[0051] Furthermore, multiple alternative particle counting measurements of the slurries in Table 1 were performed using different sensors (configured to measure particles with a particle size greater than 0.7 microns) as components of the type of particle counter shown in Figure 1. See Figure 5. In Figure 5, one measured particle size distribution curve C1 (reference) shows the expected distribution of particle sizes. Measurement curves C2, C3, and C4 were created using three different sensors and show erroneous counting at larger particle sizes. The erroneous measurements C2, C3, and C4 show overcounting of large particles and overcounted particles of different sizes. These show errors including high count (C2A) and false peak (C2B), high count (C3A) and false peak (C3B), and high count (C4).

[0052] The cause of the particle counting error, which was generated as an erroneous reading, was unclear. The applicant considered many possible causes of the error and investigated each individually in detail. For example, Figure 6 shows a decision tree that identifies potential causes of particle counting error in the system of Figure 1. Based on the numerous different causes of particle counting error that can occur in a complex optical particle counter, the possibility that the particle counting error was a result of optical feedback of laser light scattered by the slurry and returned to the laser was not considered a cause of the particle counting error.

[0053] As identified by the applicant, moving the attenuation filter to a position between the laser and the flow cell, as shown in Figure 2, eliminated the particle counting error.

Claims

1. An optical particle counter, A flow cell adapted to accommodate the flow of a liquid dispersion, A laser that generates a beam of laser light directed towards the flow cell, An attenuation filter positioned between the laser and the flow cell, configured to reduce the intensity of the laser light as it passes from the laser to the flow cell, and also to reduce the intensity of the laser light returned from the flow cell to the laser, An optical particle counter comprising an optical detection system for detecting laser light emitted from the flow cell.

2. The counter according to claim 1, wherein the filter reduces the intensity of the laser light passing through the filter by at least 50%.

3. The counter according to claim 1 or 2, comprising a photodetector for receiving the laser light that has passed through the flow cell, and adapted to detect a decrease in the intensity of the laser light caused by particles in a liquid dispersion passing through the laser light.

4. The counter according to claim 3, wherein the filter is effective in reducing mode hopping by the laser compared to an equivalent counter that does not include the attenuation filter located between the laser and the flow cell.

5. The counter according to any one of claims 1 to 4, wherein the surface of the filter is oriented at a non-perpendicular angle to the beam.

6. The counter according to any one of claims 1 to 5, wherein the flow cell is a microflow cell including a channel having a depth and width of 0.4 mm × 1.5 mm.

7. The counter according to any one of claims 1 to 6, wherein the laser is a semiconductor laser.

8. The counter according to any one of claims 1 to 7, wherein the laser light includes electromagnetic radiation having a wavelength in the range of 600 to 700 nanometers.

9. The counter according to any one of claims 1 to 8, wherein the laser has an output in the range of 7 to 15 milliwatts.

10. The counter according to any one of claims 1 to 9, comprising a liquid dispersion in the flow cell, wherein the liquid dispersion comprises a liquid medium and particles dispersed in the liquid medium.

11. The aforementioned liquid dispersion Based on the total weight of the liquid dispersion, At least 90 weight percent of a liquid medium, The counter according to claim 10, comprising less than 10 weight percent of dispersed particles.

12. The counter according to claim 10 or 11, wherein the particles are ceramic particles.

13. The counter according to any one of claims 10 to 12, wherein the particles include silicon dioxide (silica), cerium oxide (ceria), zirconia, or alumina.

14. The counter according to any one of claims 10 to 13, wherein at least 90% of the particles have a particle size of less than 0.1 microns, i.e., (D90) is less than 0.1 microns.

15. The counter according to any one of claims 10 to 14, which is adapted to detect particles larger than 0.5 microns in size by detecting a decrease in the intensity of laser light detected by the optical detection system.

16. A method for detecting particles in a chemical mechanical processing slurry using a counter according to any one of claims 1 to 15.

17. A method for detecting particles in a liquid dispersion using an optical particle counter, To provide a flow of a liquid dispersion through a flow cell, wherein the liquid dispersion includes a liquid medium having particles dispersed in the liquid medium, Using a laser to generate a beam of laser light, Passing the laser light through the flow of the liquid dispersion in the flow cell, The process of passing the laser beam through a filter before passing it through the flow of the liquid dispersion, wherein the filter reduces the intensity of the laser beam reaching the flow cell and also reduces the intensity of the laser beam returning to the laser, otherwise potentially generating optical feedback that could cause mode hopping by the laser, A method comprising detecting the laser light emitted from the flow cell.

18. The method according to claim 17, wherein the filter reduces the intensity of the laser light passing through the filter by at least 50%.

19. The method according to claim 17 or 18, wherein the optical detection system comprises a photodetector adapted to receive the laser light passing through the flow cell and to detect a decrease in the intensity of the laser light caused by particles contained in a liquid dispersion passing through the laser light.

20. The method according to claim 19, wherein the photodetector detects a decrease in the intensity of the laser light caused by particles having a size greater than 0.5 microns.

21. The method according to claim 20, wherein the filter is effective in reducing mode hopping by the laser compared to an equivalent counter that does not include the attenuation filter located between the laser and the flow cell.

22. The method according to any one of claims 17 to 21, wherein the surface of the filter is oriented at a non-perpendicular angle to the beam.

23. The method according to any one of claims 17 to 22, wherein the flow cell is a microflow cell comprising a channel having a depth and width of 0.4 mm × 1.5 mm.

24. The method according to any one of claims 17 to 23, wherein the laser is a semiconductor laser.

25. The method according to any one of claims 17 to 24, wherein the laser light includes electromagnetic radiation having a wavelength in the range of 600 to 700 nanometers.

26. The method according to any one of claims 17 to 25, wherein the laser has an output in the range of 7 to 15 milliwatts.

27. The aforementioned liquid dispersion Based on the total weight of the liquid dispersion, At least 90 weight percent of a liquid medium, The method according to any one of claims 17 to 26, comprising less than 10 weight percent of dispersed particles.

28. The method according to any one of claims 17 to 27, wherein the particles are ceramic particles.

29. The method according to any one of claims 17 to 28, wherein the particles include silicon dioxide (silica), cerium oxide (ceria), zirconia, or alumina.

30. The method according to any one of claims 17 to 29, wherein at least 90% of the particles have a particle size of less than 0.1 microns, i.e., (D90) is less than 0.1 microns.

31. The method according to any one of claims 17 to 30, wherein the surface of the filter is positioned at a non-perpendicular angle with respect to the laser beam.

32. A method for reducing laser mode hopping in an optical particle counter comprising a flow cell comprising a liquid dispersion comprising particles dispersed in a liquid, wherein the laser guides laser light into the flow cell and the liquid dispersion, a photodetector detects electromagnetic radiation leaving the flow cell, and the method comprises placing an attenuation filter between the laser and the flow cell to reduce the intensity of the laser light entering the flow cell and to reduce the intensity of the light returned to the laser as optical feedback which may cause mode hopping.

33. The method according to claim 32, wherein detecting the intensity of the beam includes measuring the decrease in the intensity of the beam as particles contained in the liquid dispersion pass through the beam.

34. The method according to claim 32, wherein the filter is effective in reducing mode hopping by the laser compared to an equivalent optical particle counter that does not include the attenuation filter located between the laser and the flow cell.

Images