$ournal o/ Appli~ Spectroscopy. Yol. 64. No. I, 1997 MEASUREMENT OF SURFACE ABSORPTION IN O P T I C A L M A T E R I A L S BY T H E M E T H O D O F L A S E R PHOTOTHERMAL RADIOMETRY O. E. Sidoryuk and L. A. Skvortsov" UDC 621.373.826:543.52 The present work is devoted to the development o/ the method o/ laser photothermal radiometry with a view to im:reas~tg its spatial resolution over the depth o/ the specunen when sat/ace layers o/ th~ substance o/ thickness about I pm or less are to be investLgated. As an example, results o/ an int,estigath~n o/ radL~tion absorption in the sur/ace layer o/ a ~th~m n ~ a t e crystal are presented. The absorption index measured at a wm,elen~ o/1.08 pm was - 0 . 6 cm -I. Key words: sur/ace absorption, laser photothermal radiometry, radiation polarization plane, Brewster angl~ The magnitude of optical absorption is one of the sensitive parameters that determine the quality of materials, specifically the presence of impurities and defects in them. This especially refers to an actual surface of a material that differs from the volume by a noticeable imperfection in the structure. The latter is generally caused by mechanical treatment of the surface [1 ] or by the presence of adsorbed molecules on it [2 ]. Naturafly, it is the surface that in many respects determines such important characteristics as laser strength, the limiting magnitude of optical losses, etc. In the majority of techniques used for measuring surface losses the magnitude of the latter is usually not normalized to the thickness of the absorbing layer [3 ]. This decreases the informative value of measurements, not allowing one to determine the existing nonuniformity of the optical properties over the specimen depth. One of the methods used for measuring extremely small magnitudes of absorption in optical materials is the method of laser photothermal radiometry (LPhR) [4-7 ]. The method is based on recording the change in the magnitude of the integral heat flux from a specimen surface exposed to irradiation and heating by periodically successive laser radiation pulses. By measuring the dependence of the magnitude of the recorded heat flux on the repetition frequency of the laser pulses, it is possible to separate the contributions to the measured signal from volume and surface absorption [6 ]. However, the LPhR method has a relatively low spatial resolution over the depth of the investigated material (SRDM); it is determined by the thickness of the layer of the substance emitting the heat flux [6 ]. Thus, for barium titanate and strontium titanate crystals, which by their properties are close to the lithium niobate crystals investigated in the present work, the thickness of the layer that is transparent within the wavelength region A ~ 10 /~m is ~ 100/~m [8 ]. Thus, a considerable thickness for the layer of the substance emitting the heat flux substantially restricts the possibilities of the LPhR method in investigating the surface of oxide dielectrics. We devoted the present work to the development of the LPhR method with a view to increasing its SRDM. One of the ways of increasing the SRDM of the LPhR method consists in using linearly polarized laser radiation incident on the surface of the investigated specimen at the Brewster angle. Then the magnitude of the surface absorption can be determined by comparing results of measurements obtained in two different cases. In the first case, we carry out measurements at a constant intensity of the laser radiation, changing the direction of its iiiii .. ii i iiii i i i i i i i i j ii ,iiiii iiiii i i iiii iiiii i|i i ii i iii i i i , ,ii [ iTi ii i [i i ] iiii i jllll ii[i i "Polyus" Research Institute, 3, Vernadskii Ave., Moscow, 117342, Russia. Translated from Zhurnal Prikladnoi Spektroskopii, Vol. 64, No. 1, pp. 82-84, January-February, 1997. Original article submitted November 21, 1995. 84 0021-9037/97 / 6401-0084518.00 91997 Plenum Publishing Corporation polarization periodically with frequency f. In the second case, we change the intensity of the laser radiation with the same frequency. First, we will consider in greater detail the former case. Suppose the initial position of the polarization plane of the laser radiation coincides with the position of the plane of incidence of the laser beam. In this case, reflection from the specimen is absent, and the radiation intensity inside the specimen is equal to the incident intensity I I under conditions of small surface, bsur, and volume, bvot, absorption. Rotation of the polarization plane of the laser radiation through the angle ~ / 2 decreases the laser radiation intensity in the specimen volume by the magnitude M (n2 -- I)2 11 = (n 2 + I)2 due to reflection losses. At the same time, in a surface layer of thickness - lO~/n (/l is the radiation wavelength, n is the refractive index of the material), where a reflected electromagnetic wave is formed [1 ], its magnitude remains constant and equal to It in any case. Thus, the recorded variable component of the thermal radiation UI is attributed only to the volume absorption bvol and is independent of the magnitude of the surface absorption bsur. As already noted, in the second variant of the measurements the intensity of the laser radiation that affects the specimen changes periodically in amplitude. Here, the positition of the polarization plane of the laser radiation incident on the specimen surface at the Brewster angle remains invariant and coincides with the position of the plane of incidence of the laser beam. It is evident that in the second case the entire specimen as a whole is subjected to periodical heating, including its volume and surface layer. Since to compare the measurement results in the two cases considered it is necessary that the change in the laser radiation intensity in the specimen volume be identical, the laser radiation intensity in the second case should change from a zero value to the magnitude 12 = (n2 - I)2 xl" (n2 + I)2 The variable component of the thermal radiation recorded in the second case can be written in the form / / 2 = U 1 +k.A, where A is the fraction of the radiation absorbed by a surface layer of thickness -lOJl/n; k is a constant of proportionality. Since for a thin surface layer we can consider that A ~-,bsur" lOb~n, then U2 -~ U 1 + kBsurlO,,I/n, whence b,,,~ = ~ n (u2 - u o . In this case we can determine the coefficient k according to [6 ] using specimens whose surface absorption is known. Figure 1 shows the scheme of measuring absorption losses in the surface layer of the material according to the above-described procedure. An yttrium aluminate cw laser l 01 = 1.08/~m) with a mean radiation power Pmean s 60 W serves as a source of linearly polarized radiation. Frequency-modulated (f = 3.2 Hz) heating of investigated specimen 6 is attained as a result of: l) the action of the radiation pulses formed by means of electromechanical chopper 2 or 2) the use of electrooptical phase changer 3, which discretely changes the position of the laser radiation polarization plane by ~ / 2 relative to the plane of radiation incidence with frequency f. A 85 3 Fig. 1. Scheme of the experimental setup: 1) laser; 2) laser radiation chopper; 3) phase changer; 4) control unit; ) generator;, 6) specimen; 7) IR objective; 8) IR detector; 9) amplifier; 10) synchronous detector. lithium niobate crystal serves as the electrooptical element of the phase changer. The heat flux from the surface of specimen 6 is focused by objective 7 onto the sensitive area of pyrodetector 8, made of a lithium tantalate crystal. Objective 7 consists of germanium lenses that are simultaneously filters protecting IR detector 8 from scattered laser radiation. After passing through amplifier 9, the electric signal from the pyrodetector is measured by synchronous detector 10, which is triggered periodically with frequency / by generator of electric pulses 5 simultaneously with phase chan~er 3 or with the stepped motor of electromechanical chopper 2. As objects of investigation, in tbi~ work we considered specimens of lithium niobate crystals that are widely used in quantum electronics, integral optics, etc. The surface of the lithium niobate crystals was previously treated by standard technology to roughness Ps " 0.032 pm and to the fourth class of purity. To eliminate the effect of reflections from the output surface on the measurement results, the specimens had a wedge shape. The required resolution of the method can be attained by using a laser radiation source of large mean power and a highly sensitive system for measuring the modulated heat fluxes. Moreover, a specific feature of the LPhR method is an increase in the signal/noise ratio in proportion to T3 with increase in the constant component of the temperature T of the specimen [6 ]. In the present work, a 1.7-fold increase in the useful signal compared to its value at room temperature was attained due to heating of the specimens to T = 800(3. The absorption coefficient at Jl = 1.08/~m measured by the above-considered method in a 5-/~m-thick surface layer of a lithium niobate crystal was equal to -0.1 cm-t. For comparison we note that the volumetric absorption coefficient in lithium niobate is - 2 . 1 0 -3 cm -! [6]. The method presented does not give us, of course, a full picture of the distribution of absorption in the surface layer of the material; however, it can be considered as one more instrument for analyzing a surface. In particular, its employment in investigations of a lithium niobate surface allows one to refine the parameters of the absorbing layer. If the minimum spatial resolution in the LPhR method with analysis of the frequency dependence of the recorded thermal-radiation signal is limited by a considerable thickness for the emitting surface layer, then it is by this resolution that the losses measured should be normalized. This gives only a tentative estimate from below for the surface absorption index. Obviously, the estimate turns out to be very rough in the case of high localization of absorption, specifically in the surface layer of lithium niobate. The high spatial resolution and good sensitivity of the LPhR variant considered in the present work allow us to consider it to be promising for investigating low absorptions in surface layers of materials. It can turn out to be useful in estimating the surface finish of products for quantum electronics and integral optics. REFERENCES 1. 86 M.S. Bakharev, L. I. Mirkin, S. A. Shesterikov, and M. A. Yumasheva, Structure and Strength of Materials under Laser Actions [in Russian ], Moscow (I 988). 0 o 4. 5. 6. 7. 8. F. F. Vol'kenshtein, Electron Processes on Semiconductor Surfaces with Chemisorption [in Russian ], Moscow (1987). M. A. Bukhshtab, Measurement of Small Optical Losses [in Russian l, Leningrad (1988). P. E. Nordal and S. O. Kanstad, Phys. Scr., 20, 659 (1979). R. Santos and L C. M. Miranda, J. Appl. Phys., 52, 4194 (1984). V. N. Lopatkin, O. E. Sidoryuk, and L. A. Skvortsov, Kvant. Elektron., 12, 339 (1985). O. E. Sidoryuk and L A. Skvortsov, Zh. Prikl. Spektrosk., 56, 781 (1992). E. M. Vomnkova, B. N. Grechushnikov, G. I. Distler, and I. P. Petrov, Optical Materials for IR Techniques [in Russian I, Moscow (1965). t 87