- Back |
- Home
- / News
- / Enterprise
- / 【 Conference Review 】 A wonderful review of the online seminar on Filmetrics light reflection film thickness gauge under KLA
重点
Introduction
Introduction
On November 19th, the online seminar "Precision Measurement, Empowering the Future: Principles and Applications of KLA's Filmetrics Light Reflective Film Thickness Gauge" jointly organized by Unicon and industry giant KLA Instruments was successfully concluded. The conference attracted numerous experts, scholars, and technical colleagues from fields such as semiconductors, advanced materials, and research institutes. A total of 244 people registered, 182 people participated deeply throughout the process, and there were up to 8296 online interactions, which confirmed the industry's high attention to precision measurement technology for thin films with tangible numbers.
In this knowledge feast, we deeply analyzed the core principles of light reflection technology and comprehensively explored its cutting-edge solutions in addressing key challenges such as measurement consistency, complex application scenarios, and efficiency improvement. This article has sorted out all the questions and answers involved in the seminar. If you want to watch the live playback, you can click the button below to jump, or follow our official account. The article or video number has a full live playback.
You can scan the code to follow the Unocal official account, or click the button to view the seminar playback.
question
&
answer, solution, explanation, resolution
Answers
Why is there a spectral reflectometer but no spectroscopic ellipsometer for measuring film thickness? Can four probes be used to measure film thickness if the top layer is made of two different metals?
At present, the KLA ellipsometer solution is still provided as an automation product for Fab customers, and desktop ellipsometers have not yet been added to the KLA instrument department.
When measuring film thickness with a four probe square resistance meter, if it is a two-layer metal film, the result obtained is a weighted average of the conductivity of the two films due to their mutual conductivity. If it is necessary to measure a single film layer on the surface, it is still recommended to add an insulation layer below the film layer to avoid mixing up
What are the core advantages of a light reflection film thickness gauge compared to an ellipsometer or other film thickness measurement techniques?
The question you raised is very good, it is actually a very core issue. Compared with other technologies such as ellipsometry, the core advantages of the light reflection film thickness gauge lie in its convenient operation, wide range, and non-destructive measurement capability. It is like a "card camera" that is very easy to operate. Users usually only need to select materials to quickly obtain stable thickness readings, which is very suitable for fast quality control and routine testing on the production line. The ellipsometer is more like a "professional DSLR" that requires fine tuning. Although it can simultaneously obtain thickness and refractive index data with higher accuracy, it requires precise parameter tuning to obtain good results, and also has higher professional requirements for operators. In terms of measurement range, optical reflectometers are also more flexible, with ellipsometers typically covering 1 nanometer to 10 micrometers, while optical reflectometers can more effectively measure thicknesses from 10 nanometers to 1 millimeter, making them adept at handling thicker films. More importantly, it is a non-contact optical method that does not damage the sample at all, which ensures the integrity of the product and allows it to be easily integrated into the production line for real-time online monitoring, which is unmatched by many sample preparation or contact based technologies such as stair treaders or scanning electron microscopes. Therefore, if your core task is to conduct fast, non-destructive, and large-scale thickness monitoring, rather than delving into the optical constants of materials, then the light reflection film thickness gauge demonstrates its irreplaceable core advantages in convenience, efficiency, and applicability.
Are there any requirements for the sample before measuring with a film thickness gauge? For example, if the surface is dirty or the sample does not have a certain degree of roughness. Does it have a significant impact on the results?
There will indeed be some impact, but the specific situation needs to be analyzed. For surface dirt, it can be regarded as an additional layer of "film". In practical operation, if the tested film layer itself is relatively thick, such as an anodized layer of about 10 microns on a mobile phone case, the nanoscale fingerprints or grease attached to it can usually have a negligible impact on the results due to the huge difference in thickness; But if measuring ultra-thin films around 100 nanometers, similar minor pollution may cause significant interference. Regarding roughness, its impact is closely related to the scale we observe: many samples that appear rough at a macro level, when measured using devices such as the Filmetrics F40 series that can provide tiny light spots of over ten microns, the small area covered by the light spot may actually be a relatively flat point, thus effectively obtaining film thickness data for many seemingly rough samples (such as heart stents, anode layers of mobile phone casings). The core principle of the light reflection film thickness gauge is to detect the optical reflection signal at the upper and lower interfaces of the film layer, as long as an effective reflection signal can be formed at the measurement point. Of course, if the overall roughness of the sample is extremely high, resulting in the inability to form stable reflections at any point, the measurement will naturally be difficult to carry out.
At present, I know that ellipsoidal film thickness measurement equipment is currently applied in the fab end. May I ask if the optical reflection film thickness measurement equipment is applied in the fab end? Thank you!
There are indeed applications. In chip manufacturing plants (Fab), the light reflection film thickness gauge has a clear and important application positioning according to different process steps. In the previous process, such as in top factories like TSMC and SMIC, for ultra-thin dielectric layers (such as tens of nanometers of silicon oxide and silicon nitride), due to their thickness being close to or even lower than the recommended ideal starting line of 30 nanometers for light reflective devices, a higher precision ellipsometer is undoubtedly a more suitable choice. However, for many film layers with thicknesses ranging from hundreds of nanometers to micrometers, light reflection devices exhibit significant advantages, with a typical application being the measurement of photoresist thickness. Some of KLA's older models, such as the OP series, already adopt a design that integrates ellipsoidal modules and light reflection modules. To better meet the measurement needs of different process segments within the Fab. In the field of post packaging and wafer level manufacturing, the application of light reflection equipment is more extensive because the film layer in this stage is usually thicker, such as the thickness of passive silicon oxide at the hundreds of nanometers, polycrystalline silicon at the micrometer level, and even auxiliary materials such as grinding wax, which can be efficiently and accurately measured by light reflection equipment.
May I ask how many nanometers can be measured by the thinnest light reflection film thickness measuring equipment? Can multi-layer transparent films be measured? How many layers can be measured at most?
Regarding the thinnest measurement limit, the device can indeed support measurements as low as 1 nanometer in terms of parameters, but this can only be achieved under extremely ideal conditions. At such a scale, optical interference is already very weak, and measurements rely entirely on reflected signals. Any small sample warping, surface contamination, or environmental interference can have a significant impact on the results, leading to extremely unstable measurement values. Therefore, from the perspective of practical application and ensuring the reliability of results, we usually recommend measuring thickness ranges starting from 30 nanometers or more.
For multi-layer film measurement, the answer is yes, the equipment has this capability. In principle, measuring multilayer films is solving a system of optical equations - with each additional layer, there is an unknown variable that needs to be solved. We have actually tested complex film systems with over ten layers, such as AR Coating. But it should be noted that the more layers there are, the more difficult it is to solve. This is mainly because a core difference between the light reflection film thickness gauge and the ellipsometer is that they usually do not directly measure the refractive index, but rely on the fixed refractive index value of the built-in material library to calculate the thickness. When there are too many layers, especially for precision optical films (whose refractive index may not be constant), solving the thickness and refractive index of all layers simultaneously based solely on reflected spectral signals will face challenges such as the non uniqueness of equation solutions, making it difficult to ensure stable measurement accuracy. Therefore, although we have cases of measuring more than ten layers, this is a special case. In conventional industrial quality control, the most recommended and reliable method is to measure film structures with three or fewer layers, such as the hard coating/permeation layer in the automotive lighting industry, or the silicon oxide/silicon nitride/silicon oxide (ONO) structure in semiconductors.
Just mentioned the material library, we often test some non-standard or new materials. What if there are no parameters in the database? Does the instrument software support us to build and fit models ourselves? Is the process more complicated?
Our system provides very flexible solutions for new or non-standard materials. If there are no ready-made parameters in the material library, you can create them yourself. This process is not complicated. You just need to prepare the refractive index (n) and extinction coefficient (k) data of the material at different wavelengths in Excel according to the format required by our software, and then import them directly into the material library of the software. Afterwards, when measuring, you can directly call the data you imported to calculate the thickness. A very practical method is that if your unit has an ellipsometer, you can first use the ellipsometer to accurately measure the refractive index data of the new material, and then import it into our system, which can achieve seamless data connection and efficient utilization.
Our software has a built-in refractive index fitting function, which is mainly suitable for common dielectric materials that conform to classical optical models such as Cauchy, such as silicon oxide, silicon nitride, and many metal oxides/nitrides. For this type of material, you can have the software perform auxiliary fitting on the refractive index. However, it should be noted that the accuracy and precision of refractive index fitting are not as good as professional ellipsometers, so we mainly use it as a reference. If the Filmetrics model is used to fit polymer materials, the accuracy of software fitting will be insufficient.
What is the minimum size of graphics that F54 can measure? Are all F54 objectives optional? Is the same hardware compatible with objectives of different magnifications?
F54 can theoretically measure 5 microns, but to ensure signal quality, it is generally recommended to measure at or above 10 microns. The fundamental reason is that the measurement relies on the reflection signal of light - the smaller the spot size, the weaker the signal intensity and the lower the signal-to-noise ratio. In addition, the actual measurable lower limit is closely related to the substrate material: it may take more than 20 microns to obtain a stable signal on glass with low reflectivity, 10 microns is usually sufficient on silicon wafers, and it may be measured to around 7 microns on high reflectivity metal substrates. Regarding objective configuration, the F54 system offers a variety of optional objectives to meet different needs, such as conventional 5x, 10x, and 50x microscope objectives with wavelengths covering 400-850nm, suitable for measuring conventional thin films such as silicon oxide and metal oxide; For ultra-thin films or specific applications, reflective objectives can also be selected, including 10x, 15x, and other specifications. All objective lenses are designed based on the same hardware platform, so the host fully supports direct replacement and compatible use of different magnification objective lenses. Users can flexibly choose according to their needs, such as using low magnification objective lenses for fast positioning of large field of view image recognition, and using high magnification objective lenses for micro area thickness measurement of small field of view. In addition to the standard model, there is also an F54-XY model in the F54 series, which is an upgraded product launched after Filmetrics was acquired by KLA. It has particularly enhanced automatic image recognition and positioning functions (Deskew and Patten Rec), making it very suitable for semiconductor chip level applications. It can automatically recognize the graphic structure on the wafer, achieve automatic chip positioning and measurement, greatly improving the efficiency and accuracy of multi-point thickness detection on wafers with repetitive graphic structures such as optoelectronic chips. Currently, it has been practically applied in enterprises such as Silan and Medikai.
Is the Cauchy model suitable for transparent films, and is light reflection also suitable for opaque films?
This is a very good question that touches upon a core concept of optical measurement. Firstly, we need to understand "transparency" and "opacity" from an optical perspective rather than visual perception. Our film thickness gauge typically operates within the wavelength range of 190-1700 nanometers, which is much wider than the visible range of 400-780 nanometers to the human eye. Therefore, many materials that appear opaque under visible light may exhibit transparency over a wider spectral range. A typical example is silicon wafers, which are completely opaque under visible light but become transparent in the near-infrared band, allowing light to penetrate and reflect back from the underlying interface, thus being measured by a reflective film thickness gauge. Therefore, the key to measurement is not whether the material "looks" transparent, but whether the tested material allows light to penetrate the film itself within a certain working band to obtain the reflected signal from the bottom interface of the film layer. As long as this condition can be met, even materials that appear opaque to the naked eye are equally applicable. On the contrary, if a material is completely opaque throughout the entire spectral range (such as certain thick metals), and light cannot penetrate to the bottom of the film layer, then the reflection method cannot measure the thickness of the thin film on it.
We have a special requirement to measure the film thickness on a curved surface. Does our equipment have this special measurement mode or accessory?
The measurement of surface film thickness on curved surfaces is a challenging application. It depends on the curvature of the curved surface? For example, in the automotive lighting industry, the hardening layer is usually plated on different surfaces of the car lights, and the surface of the car lights also has a certain curvature. In this case, using a contact probe can effectively solve the problem of signal collection. As long as sufficient effective signals can be collected, specific requirements for measuring curved surfaces can be met. At the same time, the micro area measurement mode of F40 can also solve the application problem of arc facial mask layer to a certain extent.
We often need to test dozens of samples from the same batch to evaluate uniformity. Does your instrument have an automatic sample stage? Is it complicated to set up an automatic measurement process?
If it is necessary to measure many samples from the same batch to evaluate the uniformity of the film layer, then the F50 series film thickness gauge would be a very good choice. The measurement results of a single sample will directly display the schematic diagram of the thickness of the surface facial mask layer, the maximum, minimum, average, standard deviation, thickness range, uniformity, etc.
Establishing a formula for automatic measurement is not a very complex process. Generally speaking, our engineers will directly provide a comprehensive measurement formula during the application support process. Subsequent measurements only require selecting the corresponding formula, clicking on 'start measurement', and obtaining measurement results including uniformity.
Can Si epitaxial polycrystalline silicon be measured? Is it possible to measure silicon wafers because infrared light is used?
Of course, and it's a very common application. In the semiconductor front-end silicon wafer manufacturing process, Poly Si measurement is a widely required process node.
Yes. Silicon wafers are opaque to visible light (400-850nm), but transparent to infrared light (850-1700nm), meaning that infrared light can penetrate the surface silicon film layer.
We are currently making flexible OLEDs, using soft substrates such as PET or PI, and the surface is not very flat. Can our equipment still maintain high-precision measurement in this situation? Do you have any special fixtures or measurement techniques?
We have corresponding fixtures. The flexible and uneven substrate you mentioned is indeed a common challenge in optical measurement. For such samples, we not only have corresponding solutions, but also have many successful cases in the industry. The core of your question is how to make the precise soft sample flat. The light reflection method requires the optical path to be perpendicular to the surface of the sample. Therefore, we recommend using a dedicated fixture to stably fix and flatten the flexible substrate, thereby artificially creating a local ideal measurement plane. At the same time, combined with the Filmetrics F40 series equipment mentioned by Teacher Li earlier. Using its tiny light spot to accurately locate in a micrometer level, relatively flat area for testing, and then taking statistical analysis values through multi-point measurement to ensure the representativeness and reliability of the results. Therefore, although flexible substrates bring special characteristics, high-precision measurements can be achieved through a set of methods such as "customized fixtures", "precise positioning of small light spots", and "multi-point statistics".
How does the optical film thickness gauge adapt to the environment? Will environmental vibrations or lighting affect the accuracy of measurements?
The optical film thickness gauge has good adaptability to environmental factors. In terms of ambient light, the instrument will perform background reference acquisition before measurement. Through specialized signal processing algorithms, it can effectively identify and eliminate the interference of ambient light such as daily lighting, so it can work normally without the need for darkroom conditions. In terms of anti vibration, conventional environmental vibrations have little effect on the accuracy of a single measurement, and may only cause slight fluctuations in repeatability statistics during long-term continuous measurements. Of course, if severe vibrations are encountered during the measurement process (such as hitting the workbench violently), it will still interfere with the results. Overall, the device can maintain stable operation in most laboratory and industrial environments, and in special working conditions with significant sustained vibration, configuring a professional vibration isolation table can be a reliable guarantee solution to further improve measurement stability.
How does your light reflection film thickness gauge ensure accurate coating thickness measurement in every small area of the complex three-dimensional grid structure of heart stents?
For complex structures such as heart stents, we usually use Filmetrics' F40 series, which combines its extremely small measurement spot with our specially developed positioning fixture to achieve precise measurement. The F40 series is equipped with microscopic light spots as low as 5 microns, which can accurately align each small area in the bracket grid for independent measurement, effectively avoiding data mixing distortion caused by the coverage of multiple structures at the same time due to excessive light spots. On this basis, we have developed a special measurement fixture specifically for this type of irregular device, which can stably clamp the sample and assist operators in quickly and accurately positioning the measurement spot to each designated key position, achieving efficient and controllable manual precision measurement, and systematically evaluating the uniformity of the coating on the three-dimensional structure. If you have such applications, you can contact us and communicate with technical experts for detailed solutions.
Are these pictures displayed actual measurement cases? Can we provide sample testing services?
Yes, all the displayed images are real cases of our actual measurements using Filmetrics equipment. The application range of light reflection film thickness gauge is very wide. In addition to mature applications in the semiconductor field, it can also be used in many special scenarios, such as real-time measurement of thickness changes of water droplets in the air, analysis of air box thickness, and even thickness measurement of luxury leather bag coatings. We are more than happy to provide you with sample testing support. You can choose to send the samples to our laboratory, and we will arrange professional equipment for testing and provide detailed reports; You are also welcome to visit our laboratory for technical exchange and on-site testing experience. If you have any specific needs, please feel free to contact us for further communication.
summary
As presented in this Q&A session, although the challenges of thin film measurement vary, the pursuit of precision, efficiency, and reliability is consistent. Whether your problem involves the repeatability and consistency of measurements, the analysis of complex new materials and structures, or the integration of equipment into automated production lines, solid technical expertise and rich application experience are needed as support.
If the above discussion triggers further thinking for you, or if you are facing similar measurement challenges, our application expert team is ready to provide support at any time. We are committed to translating your specific needs into reliable measurement solutions.
Welcome to initiate one-on-one technical communication with us by clicking the button or using the contact address and phone number
WeChat Official
Support
