MEASUREMENT SYSTEM AND METHOD FOR MEASURING IN THIN FILMS

Abstract

A measurement method and system are presented for in-line measurements of one or more parameters of thin films in structures progressing on a production line. First measured data and second measured data are provided from multiple measurements sites on the thin film being measured, wherein the first measured data corresponds to first type measurements from a first selected set of a relatively small number of the measurement sites, and the second measured data corresponds to second type optical measurements from a second set of significantly higher number of the measurements sites. The first measured data is processed for determining at least one value of at least one parameter of the thin film in each of the measurement sites of said first set. Such at least one parameter value is utilized for interpreting the second measured data, thereby obtaining data indicative of distribution of values of said at least one parameter within said second set of measurement sites.

Claims

1. A measurement method for in-line measurements of one or more parameters of thin films in structures progressing on a production line, the method comprising: providing first measured data and second measured data from multiple measurements sites on the thin film being measured, wherein the first measured data corresponds to first type measurements from a first selected set of a relatively small number of the measurement sites, and the second measured data corresponds to second type optical measurements from a second set of significantly higher number of the measurements sites; processing the first measured data and determining at least one value of at least one parameter of the thin film in each of the measurement sites of said first set; and utilizing said at least one parameter value for interpreting the second measured data, thereby obtaining data indicative of distribution of values of said at least one parameter within said second set of measurement sites.

2. The method of claim 1, wherein the measurement sites of said second set are selected to be substantially uniformly distributed within the thin film being measured.

3. The method of claim 1, wherein said second set of the measurement sites includes said first set of the measurement sites.

4. The method of claim 1, wherein said utilizing of the at least one parameter value for interpreting the second measured data comprises: identifying a matching set between the measurement sites in the first and second measured data; creating a temporary trained recipe for said matching set, and using said temporary trained recipe to interpret the second measured data.

5. The method of claim 1, wherein said first selected set of the measurement sites is selected to enable measurement of minimal and maximal values of said at least one parameter.

6. The method of claim 5, wherein said selecting is based on knowledge about a shape of variation of said at least one parameter of the thin film across the structure.

7. The method of claim 1, wherein said structure is a semiconductor wafer.

8. The method of claim 1, wherein the first type measurements and second type measurements are different in at least time required for collection and interpretation of measured data, the first type measurements being relatively slow measurements and the second type measurements being relatively fast measurements.

9. The method of claim 8, wherein the first type measurements are characterized by higher accuracy as compared to the second type measurements.

10. The method of any one of the preceding claims, wherein the first type measurements comprise XPS.

11. The method of claim 1, wherein the second type measurements comprise Spectral Reflectometry.

12. A measurement method for in-line measurements of one or more parameters of thin films in structures progressing on a production line, the method comprising: applying XPS measurements to a first selected set of a relatively small number of measurement sites in the thin film and providing XPS measured data; applying optical measurements to a second set of significantly higher number of the measurements sites including the measurement sites of the first set and providing optical measured data; processing the XPS measured data and determining a value for at least one parameter of the thin film in each of the measurement sites of said first set; and utilizing said at least one parameter value for interpreting the optical measured data, thereby obtaining data indicative of distribution of values of said at least one parameter within thin film, said distribution being indicative of quality of a process applied to said structure in production.

13. A system for use in the in-line measurements of one or more parameters of thin films in structures progressing on a production line, the system comprising: a first measurement device configured and operable to perform first type measurements according to a first, sparse sampling plan, and provide first measured indicative of at least one parameter of the thin film in a first selected set of a relatively small number of measurement sites in the thin film; a second optical measurement device configured and operable to perform second type measurements according to a second, dense sampling plan, and provide second measured indicative of at least one parameter of the thin film in a second set of significantly higher number of the measurements sites; and a control system configured to process the first and second measured data, the processing comprising processing the first measured data and determining at least one value of at least one parameter of the thin film in each of the measurement sites of said first set, and utilizing said at least one parameter value for interpreting the second measured data, thereby obtaining data indicative of distribution of values of said at least one parameter within said second set of measurement sites.

14. The system of claim 13, comprising a controller utility for operating the first and second measurement devices for performing said measurements according to, respectively, said sparse sampling plan and said dense sampling plan for selected measurement sites, wherein the measurement sites of said second set are selected to be substantially uniformly distributed within the thin film being measured.

15. The system of claim 14, wherein said controller operates to select said second set of the measurement sites including said first set of the measurement sites.

16. The system of claim 13, wherein said control system operates to perform said utilizing of said at least one parameter value for interpreting the second measured data by carrying out the following: identifying a matching set between the measurement sites in the first and second measured data; creating a temporary trained recipe for said matching set, and using said temporary trained recipe to interpret the second measured data.

17. The system of claim 14, wherein said controller operates to select said first set of the measurement sites to enable measurement of minimal and maximal values of said at least one parameter in said selected measurements sites.

18. The system of claim 17, wherein said controller utilizes data indicative of a shape of variation of said at least one parameter of the thin film across the structure for said selection of the first set of the measurement sites.

19. The system of claim 13, wherein the first measurement device and the second measurement device are different from one another in at least time required for collection and interpretation of measured data, the first type measurements being relatively slow measurements and the second type measurements being relatively fast measurements.

20. The system of claim 19, wherein the first type measurements performed by the first measurement device are characterized by higher accuracy as compared to the second type measurements.

21. The system of claim 13, wherein the first type measurement device is configured to perform XPS measurements.

22. The system of claim 13, wherein the second type measurement device comprises Spectral Reflectometer.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to better understand the subject matter that is disclosed herein and to exemplify how it may be carried out in practice, embodiments will now be described, by way of non-limiting example only, with reference to the accompanying drawings, in which:

[0041] FIG. 1 is a schematic illustration of the general principles of the known scheme for optical measurements, such as e.g. using OCD tool based on Spectral Reflectometry;

[0042] FIG. 2 is a schematic illustration of the general principles of the known scheme for XPS measurements;

[0043] FIGS. 3A and 3B demonstrate the effect of measurement uncertainty on the HVM;

[0044] FIGS. 4A to 4C schematically illustrate the principles of the present invention, where FIG. 4A shows a block diagram of a hybrid measurement system of the invention; FIG. 4B shows a flow diagram of a measurement scheme utilized in such system; and FIG. 4C exemplifies the arrangement of measurement sites/locations used for sparse and dense sampling plans;

[0045] FIGS. 5A to 5F exemplify measurements on a single layer film for the SiON case, where FIG. 5A shows design of experiment (DOE) set of 5 wafers measured by XPS over time; FIG. 5B shows the XPS and optical maps of measurement sites; FIG. 5C shows HESSSiON layer thickness measured by OCD tool, HESS and XPS, and N_Dose composition measured by HESS and XPS; FIGS. 5D-5E show HESS Topography maps for 5 DOE wafers, for respectively SiON layer thickness and N-Dose composition; FIG. 5F shows the precision of XPS, thickness measured by OCD tool and HESS measurements;

[0046] FIGS. 6A and 6B exemplify measurements in a multi-layer film stack, where FIG. 6A illustrates the results using standard optical measurements with two toolsets; and FIG. 6B illustrates HESS measurements with two toolsets; and

[0047] FIGS. 7A and 7B exemplify measurements on a planar stack including ultra-thin high k (HK) on top of ultra-thin interlayer (IL) dielectric layer.

DETAILED DESCRIPTION

[0048] FIGS. 1 and 2 schematically demonstrate the general principles of the known schemes for, respectively, fast (in-line) optical measurements and measured data interpretation, and slow (off-line) X-ray measurements and measured data interpretation. FIGS. 3A and 3B demonstrate the effect of measurement uncertainty on the HVM.

[0049] As indicated above, the present invention provides a thin film measurement technique, suitable for on-line measurements, particularly useful for process control of the manufacture of thin film containing structures.

[0050] Reference is made to FIGS. 4A to 4C schematically illustrating the principles of the present invention. FIG. 4A shows a block diagram of a hybrid measurement system 100 of the invention; FIG. 4B shows a flow diagram 200 of a measurement scheme utilized in such system; and FIG. 4C exemplifies the arrangement of measurement sites/locations used for sparse and dense sampling plans.

[0051] The measurement system 100 is configured to perform in-line measurements on thin-film containing structure progressing on a production line and may be aimed at process control of the production process being successively applied to the structures by a certain processing tool arrangement. More specifically, the present invention is used for process control used in the semiconductor industry and is therefore exemplified below with respect to this specific application. It should, however, be understood that the principles of the invention are not limited to this specific application. Also, the invention is particularly useful for measurement of thin film parameters, e.g. unpatterned layers, and is therefore exemplified below as being used in measuring on the uppermost thin film layer on a wafer. Further, the invention is exemplified below as using XPS and optical measurements, as respectively, slow and fast measurements. However, as mentioned above, these are non-limiting examples of the respectively relatively slow and relatively fast measurement techniques distinguishing from one another by the factors defined above.

[0052] Thus, measurement system 100 is configured as a so-called automatic measurement system in the meaning that it is associated with a production-line processing tool arrangement (e.g. is integral with such tool arrangement) to be applied to a stream of wafers progressing on the production line. The measurement system includes such main constructional parts as a first measurement device 102 configured to apply first type of measurements, being relatively slow measurements, such as XPS measurements to a measurement site on the wafer, and a second optical measurement device 104 configured to apply second type of measurements, being relatively fast measurements, such as optical measurements to said measurement site. Also provided in the system is a control system 106 which receives and processes first and second measured data MD.sub.1 and MD.sub.2 corresponding to the slow and fast measurements and generates output data indicative of the structure parameter(s), e.g. relevant for the process control.

[0053] The control system 106 may be a separate computer system connectable (via wires or wireless signal transmission) to the output of the measurement devices 102 and 104 for receiving therefrom data indicative of/corresponding to the respective measured data; or may be part of either one of the measurement devices 102 and 103; or the data processing utilities of the computer system 106 may be distributed between these measurement devices. The control system 106 thus typically includes data input and output utilities 106A, 106B; memory utility 106C; and data processor and analyzer 106D; and may also include suitable communication/data formatting utilities.

[0054] Also provide in the system 100 is a controller 108 associated with the measurement devices 102 and 104 for operating these measurement devices to implement predetermined first and second measurement plans with respect to selected locations/sites on the structure. The controller 108 may or may not be part of the control system 106. The measurement scheme 200 implemented by the system 100 utilizes the combination of the first and second measurement plans which are, respectively, so-called small sampling plan (or sparse sampling plan) 202 and large sampling plan (or dense sampling plan) 204.

[0055] The sparse sampling plan 202 is applied by the relatively slow, high-accuracy measurements, which in the present example are constituted by XPS measurements performed by the measurement device 102. Thus, N locations (generally a few locations, e.g. N=2, locations L.sub.1 and L.sub.2) on the thin film layer of the wafer W are selected, and the XPS measurements are applied to these N locations. The N locations for the XPS measurements are preferably arranged with a relatively low spatial frequency (i.e. relatively large distance between them. The XPS measured data MD.sub.1 is provided and processed to determine one or more parameters of the thin film in each of said N locations, e.g. thickness and/or certain material concentration. The so-determined parameter(s) of the thin film is/are used for learning/training mode with respect to the dense sampling plan measured data MD.sub.2 (optical data).

[0056] The dense sampling plan 204 is based on optical measurements applied to a relatively larger number M (M>>N) of measurement locations/sites on the wafer MS arranged with a relatively high spatial frequency. Preferably, the M locations of the dense sampling plan 204 include the N locations of the sparse sampling plan 202 and additional K locations, i.e. M=N+K. The dense sampling plan measured data MD.sub.2 is provided and the one or more desired parameters of the thin film determined is each of said N locations is/are used for interpreting (by fitting procedure) the optical measured data MD.sub.2 to thereby determine the distribution of one or more thin film parameters within the structure.

[0057] It should be noted that the training mode or sparse sampling plan may be applied to a few similar structure for learning the optical data interpretation. For example, the training of optical measurements using e.g. XPS measurement(s) may be applied to 5 wafers (using 9 point map in each wafer), and this measured data MD.sub.1 is used for determining the thickness and Nitrogen dose (Ndose) parameters in the uppermost layer (thin film) of the wafer. Then, the optical measurements of full wafer (e.g. 49 point map) is interpreted.

[0058] The following are some specific but not limiting examples of the measurement technique of the invention.

[0059] FIGS. 5A to 5F exemplify measurements on a single layer filmthe SiON case. FIG. 5A shows design of experiment (DOE) set of 5 wafers measured by XPS over a period of several weeks; FIG. 5B shows 5-point XPS map (N=5) and 49-point optical map (M=49); FIG. 5C shows HESSSiON layer thickness (designated as OCD, HESS and XPS) and N_Dose composition (HESS and XPS); FIGS. 5D-5E show HESS Topography maps for 5 DOE wafers, for respectively SiON layer thickness and N-Dose composition; FIG. 5F shows that XPS/XPS_OCD (HESS) precision is 10 better than non-hybrid optical measurements precision for SiON stack.

[0060] In this example, a 5-wafer DOE (design of experiment) of SiON films deposited on silicon substrate is considered. The thicknesses and N-dose are varied between wafers based on process splits. XPS is the reference toolset for SiON. In a sequence of repeat measurements over several weeks (FIG. 5A, where plot H.sub.1 represents the variability fingerprint across the wafers), the film and composition were measured, and also the slow outgassing of Nitrogen and the film growth are detected due to continued slow oxidation at the interface of dielectric/substrate.

[0061] The 9-point map data MD.sub.1 measured using the XPS device 102 was combined with a 49-point full wafer map data MD.sub.2 measured with the optical measurement device. Such map is illustrated in FIG. 5B. The results illustrated in FIG. 5C show that similar measurements using an optical technique alone (designated as OCD plot) demonstrate lower ability to identify the DOE conditions for thickness, even when the optical properties (related to N-dose) are fixed in the model due to low sensitivity. As indicated above, the term hybrid-enabled smart sampling (HESS) used herein refers to the methodology to combine/hybridize low (sparse) sampling measurements (example from reference tool) with high (dense) sampling measurements (example from workhorse tool) to allow cost effective but high performance metrology solution. The HESS methodology (HESS plot) shows substantially the same accurate measurement as the XPS method alone, for both thickness and N-dose composition.

[0062] The correspondence between the HESS and XPS graphs is further substantiated by topography maps for 5 DOE wafers shown in FIGS. 5D and 5E. The figure shows the circles representing the XPS points measured which are then extended by HESS to the whole wafer.

[0063] In addition to the striking accuracy improvement provided by HESS vs. standard optical method, the inventors have evaluated the precision improvement to a factor of 10, with HESS having similar precision to XPS (about 0.9% variability in thickness, and about 0.2% variability in N-dose). These results are showed in FIG. 5F: XPS/XPS_OCD (HESS) precision is 10 times better than non-hybrid optical measurements precision for SiON stack.

[0064] FIGS. 6A-6B exemplify measurements in a multi-layer film stack. In this example, the case of an ultra-thin oxide residue on top of a SiN and SiO2 layer stack is considered. FIG. 6A illustrates the results using standard optical measurements with two toolsets showing differences between the two tools as well as difference with respect to the reference-level XPS measurement. FIG. 6B illustrates HESS measurements with two toolsets showing good matching to XPS, correct wafer map and tool matching improved as well. It is thus shown that by applying the principles of HESS to the sparse XPS data and dense optical data, the full wafer maps are not only improved, but also the matching between the two HESS tool results and the matching to reference is improved. When reference quality data is utilized in hybrid metrology, it is possible to observe measurement performance improvement on multiple attributes, such as accuracy and matching in this example.

[0065] Reference is made to FIGS. 7A and 7B exemplifying the measurements on a planar stack including ultra-thin high k (HK) on top of ultra-thin interlayer (IL) dielectric layer. This example presents the most challenging layers within the RMG stack. Optical toolsets are typically challenged to correctly separate between the ultra-thin high-k and interlayer dielectric films (small thickness, relatively similar optical properties. This type of measurement falls well within the capability of XPS, where the photoelectron signals from SiO and HfO are not correlated. In an 8-wafer DOE, XPS identifies the DOE intent and correctly quantifies the thickness of the two layers (FIG. 7A). Using only 5 points per wafer for XPS, the HESS approach is applied using full wafer optical measurements and retrieves the similar measurement performance while reducing the measurement throughput overhead (FIG. 7B). As shown in the figures, 8-wafer DOE, XPS identifies the DOE intent and correctly quantifies the thickness of HfO.sub.2 and IL (FIG. 7A); HESS on planar target sampling (5 sites of XPS and FWM of OCD) produce similar measurement performance as that of XPS: XPS (sparse sampling sites) track HESS thickness of HfO.sub.2 and IL (FIG. 7B).

[0066] The inventors have demonstrated multiple applications (SiON, residue oxide and HK/IL) where the HESS methodology of the invention is utilized to combine sparse/limited XPS sampling with high optical across-wafer sampling to provide a cost effective and high performance metrology solution.

[0067] Thus, the present invention provides the novel approach, termed HESS approach, relating to combined or hybrid methodology where relatively slow/accurate measurement tool/device and relatively fast less accurate tool/device are used. XPS measurement device is an example of such slow/accurate tool. The XPS technique while being by itself capable of measuring the full wafer map with sufficient performance, required too much time to execute this. Therefore, considering in-line measurements applied to structures progressing on a production line, the invention uses the XPS as little as possible. A small sample set includes a few, in some examples minimum 5 points, to capture the minimal and maximal values of parameter of interest resulting from the manufacturing process. For example, if one knows that the across-wafer variation of a certain parameter of interest is typically U-shaped, then the center will be lowest point, and edges will be highest points, so the XPS measurements are applied to these points. In the case of M-shape, the center and mid-radius points are selected as the small sample set. Whatever XPS measured, this is the small sample set.

[0068] The second fast measurement tool is optical, e.g. Spectral Reflectometer. The corresponding spectra (center and edges, or center and mid-radius) are identified in the Spectral Reflectometry map (optical measured data) and hose spectra are trained against the XPS values, thereby obtaining a trained recipe.

[0069] Then, the full Spectral Reflectometry map is analyzed, and all spectra are interpreted with the trained recipe. For the small sample subset of spectra, the identical values are obtained back (as-trained), and for the other spectra, a neural network interpolation/extrapolation of the trained values is obtained.

[0070] In the above examples, small or sparse sampling plan provides a 5-point (or 9-point) map measured by the XPS, and large or dense sampling plan provides full wafer map measured by Spectral Reflectometer. The large map is analyzed to select therefrom the dies that have also a small map equivalent (for example, if XPS measures center and 4 points at the edges, then from the large map that includes all dies only center and the 4 points at the edges are selected). The optical data is interpreted with modeless recipe, e.g. train on 5, interpret on 65. In other words, this is a natural extrapolation (based on spectral data). The recipe per wafer is built. It should be noted that this method can be combined with standard modeless techniques, and provide an intermediate way where train for one wafer is used, and then verification and re-train only may be needed, or the other wafer is added to the training set as well, and so on.

[0071] It should also be noted that that the technique utilizes on-the-fly training. Indeed, small sampling plan is applied with the first measurement tool, then large sampling plan data is collected with the second measurement tool, and interpretation of data for the second measurement tool includes the following: the samples set that match between the first and second tools' measurements are identified, training is performed with respect to that sample set to create temporary trained recipe, this is used to interpret measured data from the large sample set of the second measurement tool, and the temporary trained recipe is not needed anymore. It should be understood that this technique is fundamentally different from the standard training, where DOE is designed, trained, and saves as the recipe for all other wafers. On the contrary, in the invention, this is a recipe per wafer obtained by on-the-fly training.