Open Access
Add to BibSonomy
Abstract
Modern glass producers like SCHOTT are producing and developing optical materials for various specialized applications in photonics and optical and microlithography technology. To achieve the specifications, huge efforts in metrology R&D activities have been made, and specialized tools have been developed and installed in the metrology labs over the last decades. The optical material quality requirements of such materials have been extremely high, and a continuous increase in activities has been necessary and is still required. Therefore further development and implementation of diagnostics have been performed, like optical spectroscopy and refractography, over the last decades. In the past and today, special diagnostics have been available to qualify materials for the absolute refractive index, transmission, radiation durability, and inner quality to ensure the quality of the produced materials. Methods used for this qualification are minimum deviation, precision spectral photometers, in-situ transmission measurements using UV lasers, and Rayleigh scattering. We present some past activities for metrology that have been necessary to visualize the status and the improvement of the optical quality with the help of new and continuously improved metrology.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
Corrections
31 October 2022: A typographical correction was made to the title.
1. Introduction
Since the end of the 19th century SCHOTT has produced optical glass used for special technical optical applications. The material has been used for a broad variety of different applications, like astronomy, microscopy and microlithography for example. In order to develop, produce and to qualify these materials, different approaches and steps have been obtained.
It has been found out that by control of the complete chain from development to qualification, the support by using similar or even identical characterization methods is a very efficient way. For this reason at SCHOTT diverse methods have been established, which are visualizing the required properties as early as possible within the production chain.
There a several properties and methods allowing a specific view on the current conditions of optical glass. The following different aspects have been evaluated and partially established in order to realize the required procedures:
- • Working with metrology tools for standard and highly specialized applications
- • Standardization – using ISO/EN/DE/ASTM/JOGIS - for metrology and software tools within the company and in co-operations with costumers
- • Using Standards and reference materials based on optical glass materials
- • Development of table top tools adapted to determine specific glass properties on witness samples
- • Establish an efficient and robust sample preparation
- • Implementation of data management using LIMS (Laboratory-Information-Management-Systems) and finally feed data bases and tools to implement Inline-Characterization with accuracy close to metrology lab quality for future work
With these materials the realization of microlithography systems for 365, 248, and 193 nm respectively, where a high UV photon flux (several 10mJcm−2 using ns pulsed lasers) is applied, have been enabled [1–3].
In order to control the extreme high quality specialized metrology has been set up at manufacturers qualification labs and R&D-centers. These material have to be free of point defects like color centers, which are known to be originated by raw earth elements and other critical impurities like transition metal ions. This has been made possible by improving existing process technology and the use of better purified raw materials now [1–3].
Typical specified properties are characterized for each blank and witness samples. Statistically tests are performed for initial absorption (also named internal transmission), the induced absorption (laser resistivity), auto-fluorescence, the absolute as well as the homogeneity of the refractive index and the dispersion [4].
To guarantee this large qualification amount and its quality, R&D activities within European and German projects and co-operations with costumers have been initiated. Issues in the fields of materials development, enhancement of existing metrology concerning accuracy and capacity as well as development of new metrological methods by means of fluorescence spectroscopy have been identified and in part already realized [5,6].
The past efforts to develop and qualify these materials using optical spectroscopy and refractography are presented in the following.
2. Optical spectroscopy using transmission and fluorescence methods
In older glass catalogues material data from optical spectroscopy have been shown quite rarely. For example, the transmission was taken as an integral value using broad band illumination and detection. The separation of the wavelength range have made using absorption band pass glass filters.
These filter glasses with specific broad band absorbers in glass have gained the benefit for the first optical devises to qualify the raw glass and the later glass products. A standardization has been established in the early 30is of the last century with the CIE-standards, which are partly already in use.
These devises had the drawback using film for detection as shown for the “Zeiss Schnellphotometer” (see Fig. 1). For the reading and interpretation of the obtained result, a special calibrated ruler had been used and data transfer had been performed manually.
In the following years major improvements for electrical detection had been made. Example had been the prism photometer Model DU from Zeiss Jena and the Beckmann DU, where the wavelength selection was enabled by a prism as dispersive element with 1-2 nm wavelength resolution. These new type of spectrometer had shown a significant step forward concerning quality and speed, but still far from the possibilities we see now.
When microcontroller have been available for automatic rotation of the dispersive elements (gratings and prisms) used, the modern type of one single and double beam spectrometer could be established on the market and had been widely used in universal and industrial laboratories. An example is shown in Fig. 2 for the Beckman Modell DK-2A ratio recording spectrophotometer, which had been arrived on the market in the period after 1945.
Fig. 2. “Beckman Model DK-2A Ratio Recording Spectrophotometer,” 1957, Beckman Historical Collection, Box 55, Folder 162. Science History Institute. Philadelphia [8].
When the first personal computers have arrived on the market, SCHOTT has established them into their production laboratories with beginning of the 80th of last century.
For SCHOTT as a glass producing company there has been always the challenge to adapt these type of spectrometers for their various type of optical and technical glasses, because these spectrometers have been widely used in biology, chemistry and pharmacology and not special for glasses. Therefor glass producers had to initiate further development tasks in order to get the highest possible functionality, accuracy and efficiency based on these instruments.
For example in micro lithography applications the impurities of the material are specified to be in the low ppb range. Usually the impurities in the lower ppb range have determined using analytical methods like Laser ablated Ion coupled plasma mass spectroscopy (LA ICP-MS) and Neutron Activation Analysis [9].
Analytical methods have had the drawback that they are time consuming, whereas the sensitivity for the absorption, evaluated from transmission data, have been challenging to characterize the needs (coefficient much below 10−3 cm−1). On the other hand absorption and laser resistivity of optical material are qualified with optical methods like precision spectral photometers and in-situ transmission measurements using UV lasers.
The main advantage of optical setups are the combination of high and highest sensitivity respectively and fast measurement and evaluation cycles (several minutes compared to several hours necessary for chemical analytics).
2.1 Sample preparation
The sample preparation done by grinding, polishing and cleaning have been one of the most critical and challenging processes to be considered when high precision and reliability is required.
For the optical spectroscopy the sample thickness (length of optical path) need to be prepared and characterized within +/- 0.05 mm. Moreover the planity had to be better than 1 λ, a defect free surface (acc. DIN 10110-08) and surface roughness lower than 3 nm rms for sample dimensions larger 10 mm2 are needed to fulfill the specifications. For materials used for high precise application – where a very low absorption coefficient is required like in microlithography – especially the surface roughness has to be below 1 nm rms [10].
Moreover the mechanical and chemical processes of grinding and polishing have shown partially a strong dependency on the material composition and its mechanical properties, line hardness and E-Modulus.
Historically in earlier times the sample preparation has been a manually process, which was very time consuming. During the last decades the efficiency has been improved by major improvements on tools, polishing media (slurries) and polishing processes. Examples are reported elsewhere [11,12].
The influence of polishing quality had been evaluated using results from fluorescence emission excited by 365nm (Fig. 3) for some optical glasses and SF1, which have act as a reference material at SCHOTT since the 1970ies. The emission spectra are obtained for optically polished and fine grinded surfaces. It can be seen for the other glass types that in case of grinded surfaces the data have shown a quite large variation of emission and the accuracy is much worse compared to those of the polished samples. On the other hand the emissivity for grinded samples is much higher either, mainly due to backscattering of the excitation light, which leads to a higher impact for low fluorescent glasses. The impact of a proper cleaning procedure is another aspect, which has to be considered even more carefully in daily work and needs to be evaluated very carefully and supported by diverse surface sensitive analytical methods, like White Light Interferometry or TOF-SIMS (Time-of-Flight Secondary Ion Mass Spectrometry).
Fig. 3. Investigation of low fluorescent material the excitation and emission through polished surface line identification is performed. Results have been obtained from Ref. [4].
2.2 Transmission spectroscopy
The main uncertainty factors of the photo-spectrophotometer are the wavelength accuracy, the wavelength reproducibility, the spectral resolution and the low level of stray light. The technical setup of most spectrometer producers fulfill these requirements, the quality is checked by Holmium-glass and grey filter glass standards, provided by the NIST, since the seventies of the last centuries. For details see ISO 15368 [13] for example.
Nevertheless there are differences in each setup for the fluctuation of incident flux (light source), stray light, linearity of detection system, beam divergence and baseline reproducibility, which have to be controlled very carefully and continuously in daily laboratory work.
For testing glass, it is even more important to apply an adequate sample preparation, performed by optimized grinding, polishing and cleaning procedures as described before.
When measuring the transmittance of glass samples, one has to be aware of their severe potential loss mechanisms as seen in Fig. 4.
Some of these losses are simply given by the dielectric nature of glasses, e.g. Fresnel losses and dispersion of the refractive index. Other losses are originated by non-perfect optical active surfaces, like backward scattering or surface absorption effects. These effects can overlay the volume or bulk effects, like absorption, scattering or fluorescence.
Therefore the tasks to be solved before performing the transmission measurements is to consider and ensure a proper but efficient sample preparation. During the last decades the labs of glass producers have learned to distinguish and have treated low and very low absorbing glass materials differently.
Taking Lambert-de-Beer’sches absorption law, the volume absorption coefficient kini is calculated from the measured transmission and its physical limit T0(λ), which depends on the dielectrically properties of the glass material [14]
[k_ini the spectral absorption coefficient and t the sample thickness]This limit T0 (λ) can be calculated from the Fresnel formulas and be approximated by
If n(λ) is the wavelength dependent refractive index as described in Ref. [11] already.
The required refractive index and dispersion of the glass can be obtained with a precision in the order of 10−4 applying V-Block refractometer or prism coupler (see chapter 4). For higher precision a more accurate determination of refractive index using minimum deviation principle of a prism (see 4.1) have to be applied.
As derived from Lambert-de-Beers law the efforts to lower detectable absorption coefficient, the operator have had to increase the sample thickness t. Therefore series with various thicknesses spreading at least by a factor of 20 have had been investigated and have taken in account the following technical development steps and opportunities:
- • Dynamic range of intensity (possible after implementation of electrical detectors)
- • Very thin samples (switch from manual to semi-automation preparation)
- • Very high thickness (increase machine dimension and accuracy)
Further data evaluation, using the extinction (= -log10(T)) of the measured transmittance for wavelength 600 nm is shown in Fig. 6.
One can see that by linear regression (with R2 close to 1) an absorption coefficient of 0.00004 1/mm = 0.0004 1/cm has been found. The similar results for one single sample with samples thickness of 200 mm within an accuracy (2-sigma) of 0.0002 1/cm can be performed only, when the surface roughness is better than 0.25 nm.
A typical and daily transmission measurement is shown in Fig. 7. Here the results of an internal Round Robin on a 2.03mm thick sample of B270 is seen and data for 400, 600 and 890 nm, where absorption is not expected. Therefore this have been a good prove for the accuracy of highly transmitting optical material. There measurements have been performed in four different location and partially with two different spectrometers.
At the first glance the transmission curves look very similar and similar UV-absorption edges (at T = 0,50 = 50%), located around 300 nm, have been found. Moreover no obvious absorption is seen, but below 300 nm an unexpected increase of transmission have been seen in all curves. After excluding the impact of scattering, only excitation of fluorescence light have been left for explanation of these effects and have had to be taken into account for further interpretation.
For the proof of trueness of the results obtained, the B270 samples have been send to an accredited calibration laboratory (DkD-laboratory, Hellma Optics, Mülheim), where single data points at 400, 600 and 890 nm have been investigated. In Table 1 the data at these 3 wavelength are displayed.
Table 1. results of round robin test for transmission on B270 from Fig. 7
Finally it is seen that the data converged quite well within an accuracy (2-sigma) of +/- 0,0003 = 0.3%, which has been compatible to the high accuracy given in the ISO 15368 standard.
2.3 Fluorescence spectroscopy
The characterization of glass materials on their auto-fluorescence has been started at SCHOTT at the beginning of the sixties using a Beckmann DK-2A spectrometer (Fig. 2) with filtered Xenon lamp for excitation at 365 +/- 2 nm.
These evaluations have included optical glass, filter glass and technical glass. In the beginning the observers have been interested in the optical detectable emission for qualitative classification when applying intense UV- or backlight- sources. These methods have been transferred from the well-known observations from the mineral samples [15,16].
In the seventies first standards have been established for excitation with intensive mercury line source using the prominent and intense 254 nm, 308 nm and 365 nm lines in order to obtain comparable results in each lab [17].
At SCHOTT later a fully automated Fluorolog-3 measurement system from Horiba has been installed for these measurement since 2002. The principle set-up is shown in Fig. 8. A 450 W high pressure Xenon lamp has been used as a source and with a monochromator the excitation light with an accuracy of +/- 1 nm can be chosen. The fluorescence light will emit in all directions and it is measured at an angle of 90° to the incident light. This avoids the inclusion of the incident (excitation) light in the fluorescence measurement. Afterwards the fluorescence light spectrum has been measured using a second monochromator. The operator have had to be aware to select only filters, which have no or very low autofluorescence. Data for SCHOTT filter glass have been published elsewhere based on the described methods [18].
Fig. 8. Setup of the fluorescence measurement set-up with a 45° tilted filter glass for short optical path of fluorescence light. The two monochromators are also shown, one for choosing the excitation wavelength and one for detecting the fluorescence spectrum. The Sketch of the measurement principle have been taken from HORIBA GmbH.
The accuracy (2-sigma) of the measurement procedure has been proven to be better than 7% and the reproducibility for better than 2%. This have been done using certified reference material from the BAM (Federal Institute for Materials Research and Testing, Berlin, Germany) based on dyes (BAM-F01 to BAM-F05) [19] for calibration of the detector intensity. The wavelength accuracy have been obtained by the new certified reference multi emitter glass BAM-F012 (see Ref. [20]). This work has been performed within the BMBF founded R&D project FLUOPLEX (FK 13-N8849).
The origin of the fluorescence in the past has been impurities of the raw materials used. It could be shown that with decreasing of the impurity level of raw earth and transition metal element the auto-fluorescence level has been reduced also. Moreover the redox-state and the resulting charge numbers, especially of the transmission metal elements, have a major impact on the fluorescence detected. For example the emission from ions of Mn2+, Mn4+, Pb2+, Sn2+, Cu+ and Ti4 + show a strong emission when excite them with intensive UV-light [12,13]. One can observe a lot of similarities to the fluorescence of minerals, doped crystals and organic complexes or Dyes [21,22].
Fluorescence spectroscopy enables the characterization of the elemental composition from the percent range down to the ppb-level in crystals, as fluorescence emission is known to be detectable with very efficiency [5,21, 22].
With the setup described above one can obtain emission spectra as well as excitation spectra. The emission spectra are taken by keeping the excitation wavelength constant, whereas the excitation spectra are recorded by keeping the emission wavelength constant.
Generally the absorption in the UV range is higher due to the effective absorption cross section of traces or dopants e.g. iron. As reported in Ref. [5] a comparison between optical transmission and fluorescence has been made for technical glasses. It has been concluded that the transmission barely show a small absorption around 380nm (originated by Fe3 + ions), where the fluorescent emission is visible with better signal-to-noise ratio.
Based on the described basic principles a qualitative as well as a quantitative analysis tool has been developed and proven especially for microscopy and microlithography applications. Here investigations have been made for optical glass and CaF2 and reported elsewhere [5]. The observed characteristic has been to be originated by trivalent raw earth elements (REE) induced color centers when excited by high intense ArF laser irradiation at 193 nm [23,24] because REE are known to give very intense emission due to the high absorption cross section and the high quantum yield [25–27]. Impure material would not be useable for micro lithography applications [1,23]. It has been shown that an impurity level above 10 ppb is necessary to get measurable absorption and emission. With our analytics we couldn’t find any REE traces above 1 ppb. It has to be pointed out that production level material shows no detectable fluorescence at all.
For quality control of optical glass – which has a high impact on impurity, but still below 1 ppm, the emission spectra using 365 nm excitation wavelength (known for a long time because 365 nm emission line of Mercury lamps has been widely used for different applications) have been investigated and established. Spectra for different glasses respectively quality are shown in Fig. 9. The emission bands have several maxima at 435, 485, 525 and 575 and 670 nm depending on the material composition of each glass type. These features are mainly driven by the PbO composition or substitutes (Sb2O3, As2O3, ZnO) as reported in [4] in the very low ppm-range.
Fig. 9. emission spectra of low fluorescent optical glass using 365 nm excitation wavelength. Results have been obtained in Ref. [4] already.
The evaluation of the emission spectra is performed as follows. The area integral between 400 nm and 700 nm is calculated and compared with the same area integral of the reference sample SF1. The calculated number ranges between 0 and 100%. Depending on this value the material is classified for its specific application. Typical values for the area integral of low fluorescent glasses are well below 5%. According to costumers experience this correlated well to long living and stable material during the applications. The evaluated quality can only be guaranteed when the setup arrangement is exactly identical for each measurement.
A similar type of evaluation is performed within the Japanese Optical Glass Industrial Standard JOGIS procedure which is used by Japanese companies [17] and have been valid for optical glasses also.
Since 2005 a new material type have been arrived in the optical industry, which has contained partially opposing properties like very high absorption and quantum efficiency. These Converter materials based on YAG-polycrystalline materials have had to be handled in different ways, because additional scattering as an intrinsic material property have to be carefully characterize by optical spectroscopy methods. These methods have taken into account the required high optical absorption and very high relative quantum efficiency above 80% [28–31] Therefor a governmental funded project for qualification on these converter materials using blue lasers and LEDs have been started in Germany.
2.3 Quantitative measurements
Quantitative measurement can be performed using certified standards.
Standards are commercially available at different companies or governmental institutions NIST (National Institute of Standard and Technology), PTB (Physikalisch Technische Bundesanstalt) and BAM (Bundesanstalt für Materialforschung und –prüfung), Berlin). Intense work had been performed on doped and chemically analyzed samples, based on standards which are again certificated by NIST, NPL (National Physical Laboratory) or BAM for example.
For qualitative measurements the emission, excitation and absorption spectra of the doped crystal samples have been investigated and have been evaluated based on comparison with available standards [4,32].
In order to get samples with optimized optical properties the following specifications and properties have been fulfilled by doped crystals, optical glasses and glasses based converter materials:
- • Intensity scales linearly with the dopant concentration
- • reabsorption or concentration quenching has to be avoided
- • stable optical properties at applications condition
- • temperature range: between 10 and 40 °C
- • photo stability working with Xenon and mercury lamps and VIS laser
- • photo stability against intense UV and visible light
- • diffusion barrier against moisture and oxygen
- • mechanical form stability
- • robust handling and production procedure
- • thermal stable for temperatures up to 150 °C
Fig. 10. Characteristic of the investigated standard BAMF012/LEX-1 for spectral fluorescence in comparison to the certified single values for specific wavelength.
3. Absolute refractive index diagnostics
The refractive index is one of the key-properties of optical glass and therefore refractive index measurements are key characterization methods for process control also. The requirements for refractive index control demand fast and accurate measurement method for production control and a very high accurate measurement method for tightest requirements. For refractive index measurement two different measurement setups have been common since the beginning of the 20th. Century: The V-block refractometer for economic refractive index control with up to the fifths digit accuracy and the spectral goniometer for higher precision index measurement over a broad wavelength range from 185 nm to 2.325 µm with up to the sixth digit precision [4,33].
In the past the determination of refractive index has been obtained following in-house procedures using Abbe-refractometer and transmittive prism-based refractometer, where only a few wavelength – e.g. nd and ne – could be measured as seen in Ref. [34] for example. Since many years glass producers like SCHOTT have made some efforts to create new standards for prism based refractometer (ISO 21395-1:2020) and V-Block Refractometer (ISO/FDIS 21395-2). An overview for all methods used is shown in Table 2.
Table 2. Methods for determination of refractive index at SCHOTT
Providing Sellmeier dispersion equation data for individual measurements enables accurate interpolation of index data for arbitrary wavelengths for high precision applications [33,35].
4.1 Prism based refractometer
For higher precision accuracy of ±10·10−6 for refractive index and dispersion, the minimum deviation method have been applied. Within this method - shown in Fig. 11 - a symmetric prism is prepared and the minimum deviation angle is measured in air or nitrogen atmosphere. The procedure has been automatic also and has allowed the electronic acquisition of the data. The detectable wavelength range is 248 to 2325 nm. An upgrade of this equipment performed in 2001 gives an enhanced accuracy of ±5·10−6 with an extended wavelength range of 185 nm to 2325 nm. The new and technical improved system established in 2019 has shown an accuracy of ± (3-4)·10−6 [4].
Fig. 11. Minimum deviation angle method for highest precision measurement of the refractive index, see Ref. [4].
The requirements on the accuracy depend on the wavelength range used and the type of control. Production control requires a fast testing procedure, typically lasting not longer than one hour. For the final qualification higher accuracy at shorter wavelength requires approximately 2-3 hours. In this case a better accuracy (2-sigma) of 3 × 10−6 is needed.
The requirements on this accuracy are fulfilled by using a mechanics with an angular accuracy of much less than one arcsec. During the long-lasting development processes a couple of detectors and encoders have been tested. The final setup have allowed the reduction of scattering and therefore an accuracy better than 10−5 which has been proven in Round Robin tests with the PTB in Braunschweig in 1997 and 2002. Also comparative measurements at the NIST have been successfully proven the high accuracy of the refractive index of CaF2 for 193 nm [35].
The detectable wavelength range has been extended now from 184 up to 2325 nm in 2004 [4]. This has allowed the qualification of i-line glass, fused silica and calcium fluoride for 193 nm application wavelength. As one can see, the driving wavelengths used have been given by industrial need for some specific applications.
In the past the drivers of these innovations were the astronomy, the microscopy and the microlithography. For the identification and evaluation of the spectral lines and their line widths and profiles respectively further data evaluation is another important step. Decades ago the data acquisition has been performed using photographs and ruler to obtain the spectral lines and their positions.
Different approaches have been made using basic formulas based on Kramers-Kroenig-correlation. Depending on the locations of each laboratory the mathematical evaluations have been developed in quite different ways. For example in same laboratories the Laurent-Series have been applied, whereas in other labs the Sellmeier-Formulation has been chosen [36,37]. The history of the mathematical formulation, which have been adapted the physic perfectly is also related on the available spectral lines and amount of data points [33,36]. The increase of availability for wavelength lines and their detectability in the UV and NIR-range, especially at 365 and 1014 nm has increased the performance of data processing and accuracy in industrial laboratories.
4.2 V-block refractometer
The V-block refractometer was initially designed by J.V. Hughes and his design has been published in 1941 [38]. Details are also reported in [39].
The standard accuracy (2-sigma) is around ±3·10-5 is realized with the principle of operation as shown in Fig. 12. Since the 1990th. The wavelength change and data acquisition is done automatically. With this method a wavelength range from 365 to 1014 nm is covered and more than 20.000 samples per year have been measured. This method is used for glass development, production control and fine-annealed materials [4].
Fig. 12. V-block method for standard accuracy measurement of the refractive index as shown already in Ref. [4].
The refractive index is investigated for production and quality control. For the determination of the absolute refractive index SCHOTTSCHOTT has developed and installed a highly resolving spectrometer manufactured by SCHOTTSCHOTT scientists and technicians. The absolute refractive index is essential for the optical design of optical instruments like micro lithography steppers and microscopes. Here, an accuracy less than 10−5 is required, when the highest accuracy is needed for shorter wavelength.
4.3 Prism coupler
The prism coupler allows the robust, efficient and precise determination of refractive index and dispersion on glass materials inhabiting a large index range between 1.4 and 2.4. These type of device have been originally designed for determination of refractive index of thin films on solid samples. The working principle is the generation of total reflexion and creation if surface near optical exponential decreased optical modes in the optical thinner material [40,41], when coupling the sample onto a prism with higher refractive index. By this way a tunneling effect is created, when using the combination from prism-glass-air with n_prism > n_glass > n_air.
Lasers have been used for operation, because the polarization had to be considered for its optical design. During measurements the laser hits the prism, undergoes total reflection on its lower flat surface and is finally detected. For the SCHOTTSCHOTT prism coupler – shown in Fig. 13 - the prism is rotated. If the sample is pressed smoothly versus this lower surface and perform a so-called frustrated wave – where the spatial dimension is approximately in the wavelength applied- a part of the intensity at the detector is reduced. From the angle variation and the intensity decrease at the detector the refractive index can be calculated. This procedure is similar to an Abbe-Refractometer and is therefore the same accuracy (2-sigma) range of +/- 0.0003. When using 4 different laser sources and small spot diameter of around 1mm, the dispersion between 400 and 1600 nm can be evaluated from a reduced set of four Sellmeier-Coefficients.
4. Conclusion
It has been shown that the required diagnostic to develop, analyze and qualify the high quality of optical glass has been developed over a long time frame and is now available in a very efficient and precise manner. This has been possible using precise sample preparation, continuously increasing method development and computerization. The fulfillment had been possible in collaboration with technicians, engineers and scientists having specialized know-how on glass technology and science. The detection of different impurity levels have been shown successfully using optical spectroscopy. The working range is mainly in the UV and VIS-range and therefore sufficient according to the costumer requirement. Also the flexible use of the sample geometry enables a fast and accurate detection and qualification. The same has been demonstrated for the measurement of the refractive index. Further steps have been taken to realize the use in the NIR-range above 1550 nm and the design of more compact setups as well as of new sensors and in-line diagnostic tools.
Acknowledgement
Part of the work was funded by BMBF and BMWi respectively under project number 13N8849 (FLUOPLEX), (MNPQ-Transfer, (BAM-Vh 1380), CHOCLAB and MeLuPhosQuant (project number: 03TN0015B).
Further we like to gratefully acknowledge Dr. Resch-Genger and Dr. Katrin Hoffmann from Biophotonic Group 1.10 at BAM.
Disclosures
The authors declare that there are no conflicts of interest related to this article.
Data availability
No data were generated or analyzed in the presented research.
References
1. A. Engel, K. Knapp, B. Speit, and G. Wehrhan, Ewald Mörsen: High Quality CaF2 used for 157 nm Micro Lithography, Fab Tech 14th edition (ICG Publishing, 2001), pp. 177–182.
2. A. Engel, W. Triebel, C. Mühlig, J. Alkemper, A. Krämer, J. Kandler, K. Knapp, and E. Mörsen, “Visualization of laser damage in 157 nm material CaF2 and BaF2,” 1st. 157 nm Symposium Dana Point CA, p. 391, May 2000
3. M. Mizuguchi, H. Hosono, and H. Kawazone, “Time-resolved photoluminescence for diagnostic of photoluminescence to ArF excimer laser damage to CaF2 single crystals,” J. Opt. Soc. Am. 16(7), 1153 (1999). [CrossRef]
4. A. Engel, O. Sohr, G. Westenberger, L. Bartelmess, and R. Haspel, “Advanced industrial metrology used for qualification of high-quality optical material,” Proc. SPIE 4779, 117 (2002). [CrossRef]
5. D. Ehrt, P. Ebeling, U. Natura, U. Kohlberg, K. Naumann, and S. Ritter, “Redox equilibria and ultraviolett radiation induced defects in glasses,” International Congress on Glass, Vol. 1, invited paper (2001), pp. 84–87.
6. C. Mühlig, W. Triebel, G. Töpfer, and A. Jordanov, “CaF2 for ArF lithography: characterisation by in-situ and LIF measurements,,” CHOCLAB II final report- Optics Characterisation (2003), p. 257.
7. “Zeiss Schnellphotometer,” Carl-Zeiss-Jena, CZ 32- 616a-1, https://www.mikroskop-online.de
8. “Beckman Model DK-2A Ratio Recording Spectrophotometer,” Science History Institute, https://digital.sciencehistory.org/works/v979v3150
9. “Applications and Technology of a New ICP-MS Spectrometer,” Atomic Spectroscopy, Special Issue, Vol. 16, No. 1, Jan/Feb (1995).
10. A. Duparré and G. Notni, “Multi-type surface and thin film characterization using light scattering, scanning force microscopy and white light interferometry,” In: G.A. Al-Jumaily, Optical Metrology, SPIE Critical Revue Series. Vol. CR 72 (SPIE, 1999), pp. 213–231
11. Lehrbuch Feinoptik, Chapter 5, Polieren, ISBM 978-3-00-029486-0, Jena 2009.
12. H. Y. Jedamzik and V. Dietrich, Results of a Polishing Study for SCHOTTSCHOTT XLD Glasses, Vol. 9628, (SPIE, 2015).
13. Optics and photonics: Measurement of reflectance of plane surfaces and transmittance of plane parallel elements, ISO 15368:2021.
14. H. Bach and N. Neuroth, The Properties of Optical Glass, in SCHOTTSCHOTT Series on Glass and Glass Ceramics (Springer, 2015), p. 82–85
15. H. Pick, “Structure of trapped electron and trapped hole centers in alkali halides color centers,” in “Optical Properties of Solids” Abeles, ed. (North Holland, 1972), p. 653.
16. J. S. Stroud, “Optical absorption and color caused by selected cations in high-density, lead silicate glass,” J. Am. Cer. Soc. 54, 401–406 (1971). [CrossRef]
17. JOGIS 03-2005, “Measuring Method for Intensity of Fluorescence of Optical Glass,” Japanese Optical Glass Industrial Standards JOGIS (2007).
18. S. Reichel and R. Biertümpfel, “Characterization and measurement results of fluorescence in absorption optical filter glass,” Proc. SPIE 9626, 96260S (2015). [CrossRef]
19. BAM-F01 to BAM-F05, “Calibration Kit spectral fluorescence standards,” Federal Institute for Materials Research and Testing (BAM), (2006).
20. BAM-F012, “Glass-based Multi-emitter Fluorescence Standard,” Federal Institute for Materials Research and Testing (BAM, 2014).
21. O. S. Wolfbeis, Fluorescence Spectroscopy (Springer, 1993
22. J. R. Lakowicz, Principles of Fluorescence Spectroscopy (Springer, 1999).
23. A. H. Guenther, K. L. Lewis, D. Ristau, M. J. Soileau, and C. Stolz, Laser-induced damage in optical materials, Proc. SPIE5991, 385–393 (2005)
24. U. Natura and D. Ehrt, “Modeling of excimer laser radiation induced defect generation in fluoride phosphate glasses,” NIM in Physics Research B 174, 151–158 (2001). [CrossRef]
25. D. Möncke and D. Ehrt, “Irradiation induced defects in glasses resulting in the photoionization of polyvalent dopants,” Optical Materials 25, 425–437 (2001). [CrossRef]
26. W. Geffken, “The molar absorption of different ions in glasses,” Glastechn. berichte 35, 27 (1962).
27. B. G. Ravi and S. Ramasamy, “Status of color centers in calcium fluoride,” Int. J. Mod. Phys. B 06(17), 2809–2836 (1992). [CrossRef]
28. S. Hendy, “Light scattering in transparent glass ceramics,” Appl. Phys. Lett. 81(7), 1171–1173 (2002). [CrossRef]
29. E. Zych, C. Brecher, A. J. Wojtowicz, and H. Lingertat, “Luminescence properties of Ce activated YAG optical ceramic scintillator materials,” J. Lumin. 75(3), 193–203 (1997). [CrossRef]
30. Y. Shionoya, Phosphor Handbook (CRC Press, 1998).
31. A. Engel, M. Letz, P. Nass, E. Pawlowski, B. Hoppe, C. Ottermann, U. Resch-Genger, K. Hoffmann, T. Korb, T. Zachau, D. Enseling, and V. Rupertus, Reference based Fluorescence Characterisation of Glass Ceramics Materials for High Brightness LED Applications, (SPIE, 2007) 6142.
32. A. Engel, C. Ottermann, U. Resch-Genger, K. Hoffmann, S. Schweizer, J. Selling, J.-M. Spaeth, and V. Rupertus, Glass based Fluorescence Reference Materials used for Optical and Biophotonic Applications (SPIE, 2006), 6191.
33. P. Hartmann, Optical Glass (SPIE Press, 2014).
34. JOGIS 01-2005, “Measuring Method for Refractive Index of Optical Glass,” Japanese Optical Glass Industrial Standards JOGIS (2007).
35. J. H. Burnett, R. Gupta, and U. Griesmann, “Absolute index of refraction and its temperature dependence of calcium fluoride, barium fluoride, and strontium fluoride near 157 nm,” Proc. SPIE 4000, 1503 (2000). [CrossRef]
36. H. J. Hoffman, W. Jochs, and G. Westenberger, “A dispersion formula for optical glasses and practical implications,” Proc. SPIE 1780, 303–314 (1992).
37. U. Fotheringham, F. Petzold, S. Ritter, and W. James, “Optical glass and optical design: Otto Schott´s role in the entangled development,” Opt. Mater. Express 12, 3171–3186 (2022). [CrossRef]
38. J. Hughes, “A new precision refractometer,” 1 December 1941, Journal of Scientific Instruments (1941).
39. U. Petzold, R. Jedamzik, P. Hartmann, and S. Reichel, “V-block refractometer for monitoring the production of optical glasses,” Proc. SPIE , 9628, 962811 (2015). [CrossRef]
40. R. Th. Kersten, “The prism-film coupler as a precision instrument Part I. Accuracy and capabilities of prism couplers as instruments,” Opt. Acta: Int. J. Opt. 22(6), 503–513 (1975). [CrossRef]
41. H. Onodera, I. Awai, and J. Ikenoue, “Refractive-index measurement of bulk materials: prism coupling method,” Appl. Opt. 22(8), 1194–1197 (1983). [CrossRef]
