15. - 17. 5. 2013, Brno, Czech Republic, EU

SURFACE ROUGHNESS DETERMINATION USING LASER SCANNING CONFOCAL MICROSCOPE ZEISS LSM 700 Tomáš Bezák a, Martin Kusý b, Michal Eliáš a, Michal Kopček a a

b

Institute of Applied Informatics and Mathematics, Faculty of Materials Science and Technology in Trnava, Slovak University of Technology in Bratislava, Slovak Republic, EU, [email protected], [email protected], [email protected]

Institute of Materials Science, Faculty of Materials Science and Technology in Trnava, Slovak University of Technology in Bratislava, Slovak Republic, EU, [email protected]

Abstract Surface roughness is a critical parameter which further determines properties of materials during application. Same quantitative parameters of surfaces are therefore evaluated during material design in order to bring laboratory scale closer to practice. Standard 2D measurement techniques are commonly used in technical practice for more than 5 decades. Using a diamond tip stylus the resolution of the technique is governed by the tip diameter selected for surface analysis. Further techniques such scanning electron microscopy, atomic force microscopy has been used to determine 2D and also 3D roughness. In this paper laser scanning confocal microscopy is used to determine 3D surface roughness characteristic Ra and RSa. Laser scanning confocal microscope Zeiss LSM 700 was used to measure roughness parameters of certified roughness standards in order to determine precision of measurement. Chart of objective selection was determined. Additional adjustment of the pinhole a number of optical slices is also introduced. Keywords: Laser scanning confocal microscopy, Surface roughness, 1.

INTRODUCTION

Original design of the confocal microscope was proposed and patented by Marvin Minski in 1957. Despite of the advanced solution, the idea of the microscope left unnoticed for almost 15 years. Development of light sources along with computer technology led to wide spread interest in this technique [1]. Mostly oriented on biology and biochemistry, recent design of laser confocal microscopes allows application also in the field of materials science [2]. Laser confocal microscope is a light microscopy system which uses a laser light passing through the confocal optical path for image formation. In conventional light microscopy the image is formed simultaneously for all observed points of entire field of view. In contrast, system of a laser confocal microscope forms image pixel by pixel using photomultipliers or CCD. Confocal microscopes could be divided into a two distinct groups: point scanning confocal microscopes and spinning-discs confocal microscopes. First mentioned is based on scanning the field of view using single laser beam. Its movement in x and y direction is controlled using galvanic mirrors. Later mentioned confocal microscopes are using multi point scan of the field of view. Design employs spinning modified Nipkow discs with sufficient number of lenses and pinholes rotating at suitable speed to form a real-time confocal image. Both above mentioned approaches to confocal microscopy offers advantages but also suffers from drawbacks. High precision of point scanning confocal microscopy is balance with long time for data acquisition. High speed of data acquisition – real time confocal image – of spinning discs confocal microscopes is compromised by lower resolution [1].

15. - 17. 5. 2013, Brno, Czech Republic, EU

Laser scanning confocal microscope Zeiss LSM 700 used in this study belongs into the group of point scanning confocal microscopes. Series of Rubert certified roughness standards will be used to review the capability of this microscope to deliver reliable RSa and Ra parameters. The study will provide data on objective selection. Different pinhole adjustment as well as number of optical slices will be also involved in the study as a variable affecting the precision of roughness determination. 2.

EXPERIMENTAL DEVICE AND MATERIALS

2.1

Laser scanning confocal microscope Zeiss LSM700

Point scanning confocal microscope Zeiss LSM 700 installed in the Laboratory of Structural Analysis MTF STU is based on main frame of Zeiss Axio Imager.Z1m optical microscope further equipped with a laser scanning unit LSA 700. Precision of sample positioning is guaranteed by piezoelectric linear motors of the Prior sample stage with minimum x and y step size 10 nm. Smallest increment of z drive is 10 nm. Optical train is equipped with 5, 10, 20, Zeiss EC Epiplan Neofluar, 50 and 100x Zeiss EC Epiplan Apochromate objectives. Pinhole diameter is motorize with diameter continuously variable from approx. 500 µm down to 9 µm. Illumination is provided by 0.5mW solid-state laser generating a monochromatic light with 405nm and 639nm wavelength. In this study only 405nm light source was used. 2.2

Standards and measurement variables

In total 6 certified roughness standards with sinusoidal profile with Ra: 6, 2.975 (further used as 3), 0.997 (further used as 1), 0.502 (further used as 0.5), 0.1047 (further used as 0.1) and 0.0612 (0.06) µm were chosen to evaluate the capability of the microscope to determine the profile and surface roughness average Ra and RSa, respectively. Parameters of individual measurements are summarized in table 1. Table contains 3 variables: objective (magnification and numerical aperture), pinhole size (1 or 0.3 Airy unit) and number of slices (optimal or 4×optimal). Pinhole status is defined in Airy Units (AU), where Airy unit AU corresponds to radius of the Airy disc r(Airy) [1,2,3]

1AU

2r( Airy)

1.22

NA

, nm

(1)

Numerical aperture NA is a parameter listed in table 1 for each objective used in this study and could be expressed as

NA

(2)

n sin

where, n represents refractive index of the environment between sample and objective, and α is semi-angle of the light cone immerging from the objective. Optical slice thickness OST in confocal microscopy could be estimated using half maximum of intensity distribution FWHMaxial [3].

FWHM axial

0.88 (n

n2

NA2 )

, nm (3)

Pinhole diameter in AU and corresponding value in µm for each objective used in this study were determined by computer software ZEN2009 dedicated to control laser scanning microscope, collect and process raw data. Optimal number of slices was initially suggested by ZEN2009 and this value is indicated in table 1 as „Optimal“. In order to further explore precision of data acquisition the number of slices was increased by factor of 4,recorded in table 1 as „4×Optimal“.

15. - 17. 5. 2013, Brno, Czech Republic, EU

100× / 0.95 0.3AU / 24.3 4×Optimal / 0.105

100× / 0.95 0.3AU / 24.3 Optimal / 0.105

100× / 0.95 1AU / 79.6 4×Optimal / 0.20

100× / 0.95 1AU / 79.6 Optimal / 0.20

50× / 0.95 0.3AU / 24.3 4×Optimal / 0.105

50× / 0.95 0.3AU / 24.3 Optimal / 0.105

50× / 0.95 1AU / 79.6 4×Optimal / 0.20

50× / 0.95 1AU / 79.6 Optimal / 0.20

20× / 0.50 0.3AU / 9 4×Optimal / 0.504

20× / 0.50 0.3AU / 9 Optimal / 0.504

20× / 0.50 1AU / 30 4×Optimal / 0.788

x

x

x

x

x

x

x

x

x

x

x

0,5

x

x

x

x

size

x

x

x

x

0,1

x

x

x

x

size

x

x

x

x

x

x

x

size

x

x

x

x

6 3 Standard

20× / 0.50 1AU / 30

x

0.3AU / 9

x

No. Slices / Optical Slice Thickness

Optimal / 0.520

x

0.3AU / 9

x

Pinhole / AU / Diameter

Optimal / 1.947

Optimal / 0.788

10× / 0.25

Objective / Magnification / NA

5× / 0.13

Tab. 1 Summary of individual measurements carried out on each roughness standard, where AU is used for Airy Units, Diameter for pinhole diameter in [µm] and Optical Slice Thickness [µm].

1

x

x

0,06

x

size – raw data file exceeded 5 GB where not processed The roughness evaluation was performed using ZEN2009 built-in module allowing adjustment of band-pass filter with its low-pass and high-pass range. It was however figured out that implemented high-pass filter caused deformation of collected data. Waviness of the profile have become more pronounced and contributed to roughness values determined from the profile. It was therefore decided that roughness standards do not exhibit waviness and waviness caused by data acquisition and stitching algorithm could be neglected. With this assumption only low-pass filtering was applied and all values determined as Wa or WSa are considered equal to Ra and RSa. Surface plane of the roughness standard was fitted into plane using build in fitting function in order to compensate deviation of the plane of the standard from the ideal plane perpendicular to optical axis of the microscope. 3.

RESULTS

Measurements were performed applying parameters listed in the table 1. Figure 1 shows processed data of the surface roughness standard with Ra = 0.06 µm collected with following parameters: 0.3AU, 4×optimal number of slices and 100×/0.95 objective. Lower rectangle shown in the Fig. 1 highlights the profile roughness parameters. Upper rectangle indicates areal roughness parameters.

15. - 17. 5. 2013, Brno, Czech Republic, EU

Fig. 1 Surface roughness evaluation applied to roughness standard with nominal Ra = 0.06 µm. Profile roughness value determined during measurement was 0.062 µm. Areal roughness value RSa determined from measurement was 0.068 µm. Series of graphs shown in Fig. 2 represents results obtained from measurements where variables were changed according to table 1. Each graph refers to a particular roughness standard. Nominal roughness of the standard is indicated with dash line. Graph is composed of two sets of values. One set represents RSa values representing areal roughness determined by measurement. Second set of values documents profile roughness Ra as determined from randomly selected profile. Variables used for each measurement are also listed in graphs in fig. 2 as labels for x-axis. Depending on variables the collected data exhibits good or poor agreement with nominal roughness of measured roughness standards. In general, it could be noticed that objectives with higher magnification power provides better agreement. Standards with high nominal Ra values were determined using applied variables with good precision where deviation from nominal value is on the order of hundreds of nanometers i.e. less than 0.1 % of Ra nominal value. For lower nominal Ra values such as Ra = 0.06 µm the deviation of best achieved results is at about 10 % and higher.

15. - 17. 5. 2013, Brno, Czech Republic, EU

0.14

0.14

0.5

0.5

0.12

0.12

0.4

0.4

0.10

0.10

0.3

0.3

0.08

0.08

0.2

0.2

0.06

0.1

0.3/4/20

1/1/50

1/4/50

0.3/1/50

0.3/4/50

1/1/100

1/4/100

b)

0.75

1.00 nominal roughness 0.99

0.98

0.98

0.97

0.97

c)

d)

AU/No.Slices/Objectives

AU/No.Slices/Objective

40 RSa / µm Ra / µm

6.04

35 6.02 10

8

6

6

4

4

RSa / µm

8

RSa / µm Ra / µm

6.02

nominal roughness Ra / µm

10

6.04

6.00

6.00

5.98

5.98

5.96

5.96

5.94

5.94

Ra / µm

40 35

1/1/20

0.99

0.3/1/50

0.3/4/20

1/1/50

1/4/50

0.3/1/50

1/1/100

0.45

1/4/100

0.45

0.3/1/100

0.50

0.3/4/100

0.50

1.00

1/4/20

0.55

1.01

0.3/1/20

0.55

1.01

0.3/4/20

0.60

1.02

1/1/50

0.60

1.02

1/4/50

0.65 RSa / µm

0.65

1.03

RSa / µm Ra / µm

0.70

Ra / µm

1.03

SRa / µm Ra / µm

nominal roughness

RSa / µm

0.1

AU/No.Slices/Objective

Ra / µm

SRa / µm

a)

AU/No.Slices/Objective

0.75 0.70

nominal roughness

0.3/1/100

1/4/50

0.3/1/50

1/1/100

1/4/100

0.3/1/100

0.3/4/100

nominal roughness

0.6

RSa / µm Ra / µm

0.3/4/100

0.06

RSa / µm

0.6

Ra / µm

0.16

Ra / µm

RSa / µm Ra / µm

1/1/50

RSa / µm

0.16

AU/No.Slices/Objectives

AU/No.Slices/Objectives

1/1/20

e)

1/4/20

0.3/4/5

0.3/4/10

1/1/20

1/4/20

0.3/1/20

0.3/4/20

1/1/50

2 0.3/1/50

2

0.3/4/20

nominal roughness

f)

Fig. 2 Areal and profile roughness values determined with particular set up of the confocal microscope compared to nominal roughness of standards. Nominal roughness: a) 0,06 µm, b) 0,1 µm, c) 0,5 µm, d) 1 µm, e) 3 µm and f) 6 µm 3.

DISCUSSION AND CONCLUSIONS

The laser scanning confocal microscope is considered as flexible and versatile optical technique for surface topography characterization. Frequently is compared with stylus-profiler methods and other optical noncontact methods such as phase-shifting interferometry PSI or white light interferometry WLI [4, 5]. The study presented in this paper is focused on comparison of laser scanning confocal microscopy (point confocal microscopy) with stylus-profiler verified certified standards. Varying the parameters of measurement the series of results for each certified standard was achieved. Evaluation of results from the Fig. 2 confirmed the capability of the laser scanning microscope Zeiss LSM 700 to determine roughness with good agreement with certified values predominately with those of higher nominal Ra values. However not all of parameters

15. - 17. 5. 2013, Brno, Czech Republic, EU

Objective magnification

used for data acquisition are optimal and requires thorough selection based on experience and expected Ra or RSa values of the analyzed surfaces. Results achieved in this study allow optimized selection of objectives for Zeiss LSM 700 according to Fig. 3.

100×

50×

20×

0.1

1

10

Surface roughness Ra / µm

Fig. 3 Objective selection chart of Zeiss LSM 700 as a function of surface roughness Ra It was also noticed that for roughness standard with Ra = 0.06 µm better estimate of surface roughness is achieved with increasing the number of optical slices by factor of 4 with respect to optimal number of slices suggested by ZEN2009 software. Effectiveness of using this approach for higher Ra values was not observed. Pinhole diameter equivalent to 1AU proved precision in determining plausible roughness values comparable to 0.3 AU. Further study will focus on lower Ra values as well as on repeatability and reproducibility of measurements. Multiple scans will be performed in order to determine standard deviations of RSa values. ACKNOWLEDGEMENT Authors acknowledge financial support of grant agency KEGA under project number 011STU-4/2012 This contribution is the result of the project implementation: Center for development and application of advanced diagnostic methods in processing of metallic and non-metallic materials, ITMS:26220120014, supported by the Research & Development Operational Programme funded by the ERDF. REFERENCES [1] AR Clarke, CN Eberhardt: Microscopy Techniques for Materials Science, CRC Press, 2000 [2] R.L. Price and W.G. (Jay) Jerome (eds.): Basic Confocal Microscopy, Springer Science, 2011 [3] Wilhelm, S., Gröbler, B., Gluch, M., Heinz, H.: Confocal Laser Scannig Microscopy, Carl Zeiss Jena GmbH, 2006 [4] Vorburger, T.V., Rhee, H.-G., Renegar, T.B., Song, J.-F., Zheng, A.: Int. J. Adv. Manuf. Technol., 2007, Vol. 33, pp. 110-118 [5] Al-Nawas, B., Grötz, K.A., Götz, H., Heinrich, G., Rippin, G., Stender, E., Duschner, H., Wagner, W.: Scanning, 2001, Vol. 23, pp. 227 - 231