You are on page 1of 9

This article was downloaded by: [Michigan State University

On: 10 February 2015, At: 11:16
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Molecular Crystals and Liquid Crystals
Publication details, including instructions for authors and
subscription information:

Combined Tilt and Thickness
Measurements on Nematic Liquid
Crystal Samples
H. A. Van Sprang



Philips Research Laboratories , P.O. Box 80, 000 5600, JA
Eindhoven, The Netherlands
Published online: 24 Sep 2006.

To cite this article: H. A. Van Sprang (1991) Combined Tilt and Thickness Measurements on
Nematic Liquid Crystal Samples, Molecular Crystals and Liquid Crystals, 199:1, 19-26, DOI:
To link to this article:

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or
arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms
& Conditions of access and use can be found at

van SPRANG Downloaded by [Michigan State University] at 11:16 10 February 2015 Philips Research Laboratories. Keywords: Nematic liquid crystals. Printed in the United States of America Combined Tilt and Thickness Measurements on Nematic Liquid Crystal Samples H. test cells is presented in Table I. Unfortunately.’ For use with “classical” (non superconducting) magnets of about 2 T one requires samples of more than 10 pm thickness. In the course of time a number of experimental techniques has been developed which enable measurement of this tilt angle. The incident linear polarized light is in general elliptically polarized upon exit [497]/19 . We wanted to find a method which allows for accurate determination of tilt angles without a need to measure (in a preparatory measurement) cell thickness or capacitance of an empty cell because the sample thickness might easily change during introduction of liquid crystal. It is shown that only n. The liquid crystal sample is manufactured in such a way that no twist. The Netherlands (Received July 25. For studying the tilt in samples of thickness below 10 pm we are left with either a capacitive1 or an interferometric In both methods a strong dependency is observed of the tilt angle obtained on the sample thickness. flat substrate covered. no such technique can be generally applied to arbitrary samples.A.Mol.000 5600 JA Eindhoven. Cryst. Cryst. birefringence It is common knowledge that on many surfaces non-zero director tilt angles can be obtained for the director of a nematic liquid crystal..O.. 199. A summary of those methods applicable to normal. as all of them have their own restrictions. P. pp.. In Figure 1 a block diagram is shown of the experimental set up. 1990) A new technique is presented which enables simultaneous measurement of sample thickness and director tilt of non-twisted nematic liquid crystal cells up to about 20 pm thick. It can be rotated around an axis parallel to the substrates. Vol. the ordinary and extra ordinary refractive indices of the liquid crystal. 19-26 Reprints available directly from the publisher Photocopying permitted by license only 0 1991 Gordon and Breach Science Publishers S. 1991.and n. A. Liq. The director is in a plane perpendicular to the rotation axis of the sample. Box 80. are required with no more than three digit accuracy. The method we propose is based on the crystal rotation method4p5but it is more powerful due to the fact that we measure the optical phase shift instead of the transmission as a function of the rotation angle. splay or bend are present. The only method applicable over the whole range of tilt angles is the magnetic null method.

) and the applied voltage V are controlled by a Micro computer. The rotation of the sample (Motor Drive M. Q a quarter waveplate.1 - 0-90 0. from the sample. B a Beam expander. on empty cell.D.2 n. He-Ne indicates a He-Ne laser.H .1 co. A. 3 Requires d to be constant before and after filling 8 Crystal rotation 0-13 77-90 0. 15-75 0. meas. P a fixed polarizer. 4 1 >10 4. E l l ..S Conoscopy 0-17 73-90 0. The expression for the phase difference in terms of the angle of incidence 4 with the substrate normal and the director tilt 0 with respect to the substrate is. RA a rotating analyser and PD a photodetector.' where n2 = nzcos% + nSsin20 . >20 6 < 9" in twisted structures d PD He .Ne \ Computer Phase meter FIGURE 1 Block diagram of the experimental set-up. n..2 n.n. r q I EL Remarks >10 1 Only for non-twisted samples - 1 Requires C. The quarter wave plate transforms it into linear polarized light with a rotation of the plane of polarization equal to 6 . This value is then measured using a rotating analyser as described before in Reference 7. VAN SPRANG 20/[498] TABLE I Comparison of methods to measure tilt angles Accessible Range Accuracy Method Magneto optic null method Capacitive Downloaded by [Michigan State University] at 11:16 10 February 2015 Interferometry 0 (7 8 (") Required Quantities d (km) References 0-90 0. suited for twisted samples - 2.5 d....

98.5 24.6369 1.4971 1. First.6387 1. the phase difference 6 is measured for a scan of from -30" to 30".4974 1.4973 Y c . As can be learnt from the expression for 6 only n.6372 1.o 21.0 23.4978 1. The second step is a fit of these data to the above equation using d and 8 as unknown variables. Obviously there is a scattering of the experimental data around the fitted results. This liquid crystal has been used in all samples presented here.2"C.6384 1.4976 1.6354 20. and no are left as "known" parameters.0 20.4981 1.4980 1.6391 1.5 25.6361 1. The fit of the experimental data to the theoretical formula is performed on an IBM mainframe computer using standard IMSL Fortran software.91 p m thick cell with a Polyimid orienting layer.6380 1.-0 I m 145 - c ? 0) 3 Angle of incidence (") - FIGURE 2 Phase Retardation versus rotation angle for a 1. A typical example of the results is shown in Figure 2 for a 1.5 21 . expressed as a standard deviation of 0.5 22.6365 1. . temperature.0 24.4984 1.Downloaded by [Michigan State University] at 11:16 10 February 2015 TILT AND THICKNESS MEASUREMENTS [4991/21 It is this equation which is used directly in our procedure.5 23.4983 1.4975 1.4979 1.27".91 pm thick PI cell with 8 Measured at 22. In Figure 3 we show the results for an TABLE I1 Refractive indices for ROTN 3010 at A = 632 nm 1.0 22.0 1.6376 1. = 0. We measured the relevant indices in a slightly modified Pulfrich Refractometer which has an accuracy of better than In Table I1 we present the refractive indices for ROTN 3010 around room 1. Obviously d is directly proportional to 6 and determines the position of the 6-+ curve on an absolute scale while 0 determines the shape of the curve. This leads to an almost independent determination of d and 8 due to the fact that both variables occur in a different way in the expression for 6.6357 1.

evaporated SiOz layer which had been rubbed subsequently.3"C.13".rotation angle for a 2.16 prn thick SiO. 4 t a. cell.54" only in Figure 3 . sample with R Mcasured at 25. A. As can be seen the shape of the curve is different as is the observed deviation of the fit from the experimental data which amounts to a standard deviation of 0. b) data for cell in Figure 3.H." -2 9 25. m = -40 -30 -20 tij -10 0 10 20 Angle of incidence (") 30 40 t L e -40 -30 -20 -10 0 10 20 L w Angle of incidence (") 30 40 FIGURE 4 Comparison of the deviation of the fitted data from the experimental data as a function of thc rotation angle: a) data for cell in Figure 2.S. To understand this difference the experimental technique must be considered in more detail. -e 2 c c 0 . During the experiment the sample rotates over 60" and the laser beam (diameter 1 mm) passes through + L- . VAN SPRANG 221[500] Downloaded by [Michigan State University] at 11:16 10 February 2015 -40 0 -20 20 Angle of incidence (") 0 - FIGURE 3 Phase Retardation V. Apparently in the PI cell the deviation consists of abrupt humps as a function of while a gradual change is observed in Figure 4b for the SiO. In Figure 4 we compare the SiOz and PI cell.

(Magnification 100 x ) . cells exhibit a more gradually changing noise FIGURE 5 Micro photographs showing the structural difference between: a) PI oriented and b) rubbed SiO.TILT A N D THICKNESS MEASUREMENTS [501]/23 Downloaded by [Michigan State University] at 11:16 10 February 2015 slightly different positions at the liquid crystal-orienting layer interface. From the photographs in Figure 5 it is immediately clear that PI is much less homogeneous at a microscopic scale than is SO. . Although individual differences occur. Thus inhomogeneity is supposed to be the reason for the substantial noise in PI samples. See Color Plate I. it is generally true that S O . oriented samples. The photographs were taken under dark-field illumination and using crossed polarizers..

= 22.58 pm. As discussed before a certain inhomogeneity might lead to a higher noise level. it is necessary to know the absolute value of 6 and not the value modulo (360") as determined in the setup of Figure 1. This does not seem to greatly influence the accuracy of the fitted results.-0 I R1 %0 c L 350 400 300 I 250 -40 I I -20 I I 0 I I 20 Angle of incidence (") I 40 FIGURE 7 Phase retardation vs rotation angle for a 61. For samples much thicker than 20 p. it also has its limitations. VAN SPRANG 24/[502) as a function of d. Figures 6-8 contain further illustrations of the method. than PI samples.0"C. The number of times a 360" phase jump is encountered is registered by the computer and used to determine 6 at = 0.85". 3. The accuracy of the refractive index data is very important to obtain accurate Downloaded by [Michigan State University] at 11:16 10 February 2015 + 1- 5 c 0 .8 pm SiOz cell with 8 at 25. We can summarize them as follows: 1. 2._ % P m M ?? 480 470 460 450 - 440 - ul w m 430 I I I I I I I c v C . For cells with 6 (632 nm) > 360". A. the phase jumps might occur within a too small voltage range and hence the 6 value could be wrong by 360" or more.H. or equivalently d8n > 1. Although the method is powerful. For this reason we use samples supplied with I T 0 electrodes to which an increasing voltage up to 1OV is applied. Measured .

13 2. Thus in general the same conclusions hold independent of tilt or thickness.498 1.82 6.85 22.636 1.517 6 (“1 22.635 1.77 6.497 1.635 1..on the obtained 0 and d values n. From Tables IV and V similar results can be extracted for the data displayed previously in Figures 4 and 6 respectively.634 1.85 22.87 22.496 1.1 pm.6351 1.4 6.634 1.47”.c 800 m E a c s? % ! m .4973 1.496 1. 110 1.636 1.636 1.635 1. In Table I11 we show how n.16 d(w) 2.497 1.15 25.14 d (l.1 deg while the thickness is more accurate than 0.85 22.c a 600 400 I -40 I I -20 I I I 0 I 20 Downloaded by [Michigan State University] at 11:16 10 February 2015 Angle of incidence (”) I 1 40 - FIGURE 8 Phase Retardation vs rotation angle for a 14. n0 1.635 1.83 22.16 2. and no for a thin sample n. TABLE 111 Influence of the variation of n. and no within 1 from their exact value an accuracy in e is still obtained of better than 0.655 1. Concluding we can say that the method is apparently accurate enough for appli- .59 pm SiOz cell with 0 = 30.13 25.498 f3 (“1 25.72 6.10 25.TILT AND THICKNESS MEASUREMENTS [503]/25 1200 1000 C 0 .87 22. Measured at 2 I .82 6.92 6.84 TABLE IV Variation of n.497 1.84 22.498 1.82 23.4972 1.498 1.13 25.87 6.19 8 and d values from the fit.634 1.496 1.497 1. In general this allows for the conclusion that for values of n.4”C. and no influence the results for the data of Figure 7. and n.6353 1.77 6.636 1.87 6.16 2.16 2.

Kosmowski.89 cation to experimental samples without twist having a tilt between 0-90" degrees and a thickness up to about 20 pm. D. 253 (1984). 421 (1983). 104. Displays April 1984. 9 . B. J . Res. W.499 1.. Scheffer and J.. Appl. Simon and D. J .50 1. van Sprang. Phys. Inhomogeneity of the orienting layer leads to noise in the experimental data which does not significantly influence the results of the fit. Physique.81 5. Nicholas. Techn. 48.33 5. Komitow. D. H. 5 (1977) and references therein. 2001 (1976). T.637 1. Res. 18.98 5. 4.89 5. Morrissy and B. E. 3.. Hawk and H.6380 1. D. 1423 (1985). A. Becker and D. A. Phys. 2. 23.H. A. H... Zmija. Nehring. J . Opara.64 1. B. . W. Baran and J. J. References 1. Techn.497 1.639 1. M. 19.32 1. J. Koswig.4980 1. Needham. Cryst. 44. Downloaded by [Michigan State University] at 11:16 10 February 2015 1. J . R. Phys. 6. A. Mlynski. G . The method described above yields a reliable way to determine both tilt angle and thickness in cells which are made under normal TN processing conditions but without a twisted director pattern. 1073 (1988). A. J. Crossland. 5. and n. VAN SPRANG 26/[5041 TABLE V Variation of 0 and d with n. Crysr. T.33 1.33 1. 7.. M.