US 20060255296 A1
A method that is sensitive to lattice damage (also called “primary method”) is combined with an additional method that independently measures one of two parameters to which the primary method is sensitive namely dose and energy. In some embodiments, the additional method is sensitive to dose, and in two such embodiments 4PP and SIMS are respectively used to measure dose (independent of energy). In other embodiments, the additional method is sensitive to energy, and in one such embodiment SIMS is used to measure energy (independent of dose). Use of such an additional method resolves an ambiguity in a prior art measurement by the primary method alone. The two methods are used in combination in some embodiments, to determine adjustments needed to match two or more ion implanters to one another or to a reference ion implanter or to a computer model.
1. A method of adjusting an ion implanter, the method comprising:
evaluating a first wafer using a first method that is sensitive to a first parameter selected from a group consisting of dose and energy, the first method measuring the first parameter independent of a second parameter in said group;
determining a first adjustment in the first parameter based at least on a first measurement obtained from said evaluating of first wafer;
evaluating a second wafer using a second method that is sensitive to both the first parameter and the second parameter, wherein the second wafer is implanted by said ion implanter after said first adjustment is made; and
determining a second adjustment in the second parameter based at least on a second measurement obtained from said evaluating of the second wafer.
2. The method of
the first parameter is dose and the second parameter is energy.
3. The method of
the first method uses a four point probe (4PP) apparatus to measure dose during said evaluating.
4. The method of
the first method uses secondary ion mass spectrometry (SIMS) to measure dose during said evaluating.
5. The method of
the second method uses a carrier wave.
6. The method of
the second method uses carrier illumination.
7. The method of
using a first ion implanter to implant a plurality of reference wafers at a corresponding plurality of doses;
evaluating the reference wafers using the first method to obtain calibration data for the first method;
evaluating the reference wafers using the second method to obtain calibration data for the second method;
before said evaluating of the first wafer, implanting the first wafer by using the second ion implanter set to said predetermined value of the first parameter;
wherein said determining of first adjustment comprises using the first measurement with the calibration data for the first method to determine a first dose, and comparing said first dose with said first measurement to identify said first adjustment;
making the first adjustment on the second ion implanter; and
before said evaluating of the second wafer, implanting the second wafer with the second ion implanter after said first adjustment is made;
wherein said determining of second adjustment comprises using the second measurement with the calibration data for the second method to identify said second adjustment.
8. The method of
the first parameter is energy and the second parameter is dose.
9. The method of
the first method uses secondary ion mass spectrometry (SIMS) to measure energy during said evaluating.
10. The method of
the second method uses a carrier wave.
11. The method of
the second method uses carrier illumination.
12. The method of
implanting the first wafer with the ion implanter before said evaluating of the first wafer;
making the first adjustment on the ion implanter; and
implanting the second wafer with the ion implanter after said first adjustment is made.
13. The method of
the first method uses a dose measurement technique during said evaluating.
14. The method of
using said ion implanter at nominal dose and nominal energy to implant a third wafer;
storing the third wafer unprocessed while implanting a plurality of wafers using said ion implanter at nominal dose and nominal energy and further processing said plurality of wafers;
wherein said evaluatings and determinings are performed subsequent to said storing; and
wherein at least one of said determinings uses calibration data from said third wafer.
15. A method of adjusting an ion implanter, the method comprising:
evaluating a first wafer using a dose-sensitive method;
determining an adjustment in dose based on (a) a measurement obtained from said evaluating of first wafer, (b) calibration data for the dose-sensitive method, and (c) a dose used by an ion implanter in implanting the first wafer;
evaluating a second wafer using another method that is sensitive to both dose and energy, wherein the second wafer is implanted by said ion implanter after said adjustment in dose is made; and
determining an adjustment in energy for said ion implanter based at least on another measurement obtained from said evaluating of second wafer.
16. A system for matching a plurality of ion implanters to one another, the system comprising:
a first implant metrology tool sensitive to dose of implants in a first wafer;
a computer, coupled to the first implant metrology tool, the computer being programmed to determine a dose adjustment, based at least on (a) a first measurement of dose from the first implant metrology tool and (b) a predetermined value of dose in an ion implanter used to implant the first wafer;
a second implant metrology tool sensitive to dose and energy of implants;
wherein the computer is further coupled to the second implant metrology tool, and the computer is further programmed to determine another adjustment to said ion implanter, based at least on (a) a second measurement obtained from evaluation of a second wafer by the second implant metrology tool and (b) a predetermined value of energy used in said ion implanter to implant the second wafer after said dose adjustment.
This application is a continuation of U.S. application Ser. No. 10/779,072 filed Feb. 13, 2004 which is incorporated by reference herein in its entirety.
In the processing of a semiconductor wafer to form integrated circuits, charged atoms or molecules (such as arsenic, boron or phosphorous) are directly introduced into the wafer to create doped regions, in a process called ion implantation.
Ion implantation normally causes damage to the lattice structure of the wafer. Specifically, in the implantation process, silicon atoms are knocked out of the lattice, and result in vacancies. To remove such lattice damage, the wafer is normally annealed at an elevated temperature, typically 600° C. to 1100° C. Prior to annealing, material properties of the doped regions (such as the concentration of the implants and the depth of the implants) may be measured.
In an implant metrology tool, secondary ion mass spectrometry (SIMS) can be used to determine the profile of the ion implants (i.e. their concentration as a function of depth from the wafer surface). The depth of the profile (at a nominal dose) indicates a measure of the energy used in the ion implantation process. The area under the profile is a measure of the dose of the ion implants.
The SIMS process involves bombarding the wafer with atoms (e.g., oxygen atoms) which collide with atoms on the wafer surface and cause the atoms to be ejected from the surface (i.e., sputtering). During the energy transfer process, a small fraction of the ejected atoms leave as either positively or negatively charged ions which are collected by the mass spectrometer. The ion yield of the wafer is measured and a linear dependence between the ion yield and the doping concentration is used to determine the profile. Since material is sputtered form the processed wafer, causing damage to the sputtered region and re-deposition on nearby regions, the test is destructive and cannot be performed on production wafers during semiconductor wafer fabrication.
Another implant metrology tool uses a four point probe (4PP) arrangement. Specifically, the 4PP arrangement contacts the wafer's surface with four probes arranged in a straight line. Two outer-most probes are used to establish a flow of current through the implanted layer, while the two inner probes are used to measure a voltage drop. The measured voltage is used to deduce the sheet resistance and hence conductivity of the implanted layer and hence the dose. Note that the 4PP arrangement is insensitive to energy used in the ion implantation process. The probes must make contact with the wafer surface, which is considered a destructive process. Also, the measurement needs a large open area of several square millimeters. For these reasons, the 4PP technique is not normally used for process control on production wafers during semiconductor wafer fabrication.
There are several implant metrology tools that measure the damage to the lattice structure in a non-contact manner which is essential to monitor and control the fabrication of wafers. For example, a brochure entitled “TP-500: The next generation ion implant monitor” dated April, 1996 published by Therma-Wave, Inc., 1250 Reliance Way, Fremont, Calif. 94539, describes an implant metrology tool called “TP-500” that requires “no post-implant processing” (column 1, lines 6-7, page 2) and that “measures lattice damage” (column 2, line 32, page 2). The TP-500 includes “[t]wo low-power lasers [that] provide a modulated reflectance signal that measures the subsurface damage to the silicon lattice created by implantation. As the dose increases, so does the damage and the strength of the TW signal. This non-contact technique has no harmful effect on production wafers” (columns 1 and 2 on page 2). Such a TW signal is believed to be the result of carrier waves as described in one or more of the following U.S. patents all of which are incorporated by reference herein in their entirety, namely U.S. Pat. No. 5,042,952 granted to Opsal et al., US patent.
Damage to the lattice structure can also be evaluated in a non-contact manner by illumination of carriers without creating waves by use of another implant metrology tool called “BX-10” that is available from Applied Materials, Inc., 3050 Bowers Avenue, Santa Clara, Calif. 95054. Carrier illumination is briefly described in U.S. Pat. No. 6,656,749 granted to Paton et al. which is incorporated by reference herein in its entirety. This patent suggests detecting the depth of the source/drain region using a low power laser to excite carriers in the active silicon and a second laser to illuminate the surface. Through use of interferometry, the difference in index of refraction between silicon with excited carriers and silicon with non-excited carriers is to be determined. From the difference in index of refraction, a measurement as to the depth of the activated source/drain regions is to be made. Carrier illumination methods may also be performed as described in each of the following U.S. patents all of which are incorporated by reference herein in their entirety, namely U.S. Pat. No. 6,489,801, U.S. Pat. No. 6,426,644, and U.S. Pat. No. 6,483,594 and U.S. Pat. No. 6,323,951.
The above-described non-contact methods appear (to the inventor of the current invention) to have the following drawback. Specifically, to the extent such methods are based on measuring the damage to the lattice (i.e. the vacancy concentration), the inventor notes that the damage is a function of both dose and energy used during ion implantation. The inventor further notes that higher energy or higher dose, each creates a greater number of vacancies. For this reason, a change in a signal measured by the lattice-damage sensitive methods can indicate a change in either energy or dose (without being able to distinguish therebetween). Hence, an error in dose is easily confused with an error in energy when using a lattice-damage sensitive method.
U.S. Pat. No. 5,862,054 which is incorporated by reference herein in its entirety describes a method to monitor process parameters from multiple process machines to provide real time statistical process control. In this patent, the particular implementation was derived from ion implantation of wafers, but has wide applicability where there are a number of process machines having a number of process parameters and close continuous sampling of data is required. The process parameters are collected on a single computer over a single RS 485 network, and each parameters is analyzed and displayed separately for each process and process machine. Statistical variables like Cp and Cpk arc calculated and presented on the computer screen along with graphs of the various parameters for a particular process machine. Data is aged out of the computer to an archival data base under the control of a manufacturing information system and connected to a company wide network.
U.S. Pat. No. 6,408,220 which is incorporated by reference herein in its entirety describes a a manufacturing environment for a wafer fab, and a statistical process control (SPC) environment for setting control limits and acquiring metrology data of production runs. A computation environment processes the SPC data, which are then analyzed in an analysis environment. A manufacturing execution system (MES) environment evaluates the analysis and automatically executes a process intervention if the process is outside the control limits.
In accordance with the invention, a method (also called “primary method”) that is inherently sensitive to two parameters of a wafer fabrication process, is used with an additional method that independently measures at least one of the two parameters. Specifically, in some embodiments, measurements made by the primary method and the additional method are used together to match to one another multiple apparatuses each of which implements the wafer fabrication process, so that all apparatuses perform the wafer fabrication process in an identical manner. After matching, wafers produced by the multiple apparatuses have identical values for the two parameters, regardless of the apparatus that was used.
In several embodiments of the invention, the apparatuses are ion implanters and the primary method is sensitive to lattice damage, and hence it is inherently sensitive to both dose and energy. In some embodiments, the additional method is sensitive to dose of implants and in one such embodiment a four point probe (4PP) apparatus of the type described above is used to measure dose (independent of energy). Use of the additional method resolves an ambiguity in the prior art measurement by use of the primary method alone, which as noted above is inherently sensitive to both parameters (dose and energy).
The two methods are used in combination to determine adjustments needed to match two or more ion implanters as follows. Specifically, in some embodiments, a first wafer that has been implanted by an ion implanter is evaluated using the additional method, to obtain a measurement indicative of dose. Next, an adjustment to be made in the dose is determined based at least on (a) the just-described measurement and (b) a predetermined value for the dose as used in implanting the first wafer. Then the dose on the ion implanter is changed by the just-described adjustment.
Next, a second wafer is implanted by the ion implanter which has been adjusted as described above. Then, the second wafer is evaluated using the primary method (which as noted above is sensitive to dose and energy). Then an adjustment in energy is determined based at least on (a) a second measurement obtained from evaluation of the second wafer by the primary method and (b) a predetermined value of energy used in the ion implanter when implanting the second wafer. Then the energy setting in the ion implanter is changed by the just-described adjustment.
The above-described process may be repeated for any number of ion implanters, to match several ion implanters to one another (because all such implanters are matched to predetermined values of dose and energy). Moreover, the primary method and the additional method may be calibrated on an ion implanter which acts as a reference (in which case all other ion implanters are adjusted to the reference implanter).
A method that is sensitive to lattice damage (also called “primary method”) is used, in accordance with the invention, with an additional method that independently measures either dose or energy. In some embodiments, the additional method is sensitive to dose, and in two such embodiments 4PP and SIMS are respectively used to independently measure the dose. In other embodiments, the additional method is sensitive to energy, and in one such embodiment SIMS is used to independently measure energy. Regardless of which parameter (dose or energy) is measured by the additional method, use of the additional method resolves an ambiguity in use of primary method alone (due to its sensitivity to both dose and energy).
In one particular embodiment, a computer 200 (
Initially, an ion implanter 211 is designated as a reference. A set of wafers 230 (say five wafers) are thereafter implanted (as per act 401 in
Next, in the just-described embodiment shown in
The measurement from the dose-sensitive method is then used with above-described data (also called “calibration” data) from reference wafers (for the dose-sensitive method) to look up the dose (this looked-up dose is now in accordance with the reference implanter 211). Specifically, a measurement 501 from tool 221 is illustrated in
Such a lookup may be performed manually in some embodiments, although it may also be performed by a computer. In the embodiment illustrated in
Next, as per act 405 in
At this stage, a dose knob on the second implanter 212 has been offset from the nominal dose (as indicated on the second implanter) by the just-described difference amount. In the above-described example, since the looked up dose of 0.85×1015 atoms/cm2 is lower than the nominal dose 1.0×1015 atoms/cm2, the dose knob of the second implanter 212 is now set at an increased setting, i.e. now appears to be implanting 1.15×1015 atoms/cm2 (which is greater than the nominal dose by the difference amount). However, in actual fact both implanters 211 and 212 hereinafter impart the same identical dose (i.e. the nominal dose as per the reference implanter) to all wafers processed therein.
Next another wafer 232 (also called “energy-calibration wafer”) is implanted as per act 407 in
Specifically, when the measurement 511 is used with the dose line 513 for tool 222, as shown in
Assume that the difference between values 511 and 512 is 2000 uV, then this difference may be multiplied by the slope of the energy curve (which is shown in
Such additional calibration data may, for example, relate measurements from tool 222 to corresponding energy levels (not shown, but the graph would be similar to
Next, as shown by act 410 in
Such a method may be repeated if there are any additional ion implanters (such as implanter 213) that need to be matched to the reference ion implanter 211, as illustrated by the “no” branch out of act 411 in
Note that, the just-described method may be repeated after waiting for some time (as per the “yes” branch of act 411 and act 412) to re-align implanters 211-213 to one another, for example in case they drift away during wafer fabrication, or in case maintenance work was done on an implanter. The duration of waiting in act 412 depends on specific characteristics of the ion implanters and is determined by trial and error. Such a waiting period may be based either on a fixed time interval or on the number of wafers that have been processed or any other appropriate measure of usage of ion implanters.
In one specific embodiment, the energy resolution (
In an embodiment that performs the method of
In some embodiments, one or more wafers that are implanted by the reference implanter 211 at the time it is originally calibrated (for a given implant process) are stored away unprocessed in order to provide a standard for future calibrations. For example, one of the above-described reference wafers which already has the nominal dose of 1.0×1015 atoms/cm2 and nominal energy of 2 keV for the current implant process may be kept unprocessed (and is hereinafter referred to as a “golden wafer”). If such a reference wafer (at the nominal energy and nominal dose) doesn't exist, then it is created on reference implanter 211. Thereafter, the matched ion implanters (including the reference implanter) are used in the normal manner to process wafers (called “production wafers”) while the golden wafer remains unprocessed. The golden wafer is saved away to provide a snapshot of dose and energy settings within the reference implanter 211 at the time of original matching of ion implanters for the current implant process (e.g. when act 402 of
At any time in future, such a golden wafer may be used in the above-described manner to adjust one or more ion implanters back to their original values, e.g. to perform act 403 directly after act 412 (
Creation and use of golden wafers ensures that the current implant process is repeatably performed at any time in future in an identical manner to the past, even if the reference ion implanter itself drifts from its original settings, is used for a different implant process in the interim, or loses its original calibration because of maintenance work. Re-creation of a current process at any time in future becomes possible because even the reference ion implanter may be appropriately adjusted for dose and energy in embodiments of the type described above, simply by use of a golden wafer.
Therefore, in some embodiments, one or more golden wafers are used to match only the reference implanter to its original settings (i.e. the entire matching method is performed only on a single ion implanter, namely the reference ion implanter). In such embodiments, act 411 (
Note that such golden wafer(s) (which contain a snapshot of dose and energy implanted therein) can also be used to re-create a current process at any wafer fab, at any location (i.e. anywhere in the world) because such golden wafer(s) can be quickly and easily transported to any location. Thus use of golden wafers allows flexible performance of a current process at any geographic location and also at any point in time.
Identity between dose and energy values in the past and in the future are not required in some embodiments, e.g. if a current process will never need to be performed in future, e.g. if a new version of the current process is to be performed. In such a case, “golden wafers” are not required to be kept across multiple version of the process. Instead, after act 412, the entire process is repeated by returning to act 401 (i.e. instead of returning to act 403, reference wafers are again prepared in act 401, in order to match ion implanters to one another once again).
Although one or more golden wafers are described above as being prepared at nominal dose and nominal energy, other embodiments may use golden wafers at dose and energy levels that are different from the nominal dose and nominal energy, e.g. at one or more predetermined sets of dose and energy levels that are within upper and lower limits for a given process (e.g. approximately in the middle).
Note that in
Numerous such modifications and adaptations of the embodiments and examples described herein are encompassed by the attached claims.