Question: Answer: Question: Answer: Question: Answer: Question: Answer: Question: Answer: Question: Answer: COMMENTS Were these photoacoustic measurements all done on the same site? Yes. Okay, you did not see any conditioning effect with a reduction of the acoustic signal as you proceeded on a particular site under the same conditions? No. This effect has been observed on wide gap materials recently at short wavelengths. Not in this case. Thank you. Did you do the measurement where you turned off the probe pulse and just measured the acoustic signal through the pump alone? Yes we did. It shows that with sodium chloride the slope is two and with silicon dioxide it is three, so we conclude that in sodium chloride you have basically two photon absorption and in silicon dioxide you have three photon absorption. I guess what I would have liked to have seen is a curve like shown with a photo acoustic signal, but just with the probe. I mean just with the pump. You When we do an experiment we turn the pump in the pulse down to the minimum. cannot see a signal on the display. Then we increase the heating pulse on so that we see a signal. The experiment is done under the condition of very weak pump signal due to pump pulse. How is the temperature calculated? The temperature is based on equating the increase in lattice temperature to the free electron heating mechanism. It is related to the ratio of R to P, which is obtained by fitting the data. This is the equation we used to calculate temperature and you see that for arbitrary e, you can calculate any temperature at any photon energy. How do you expect the heating results to scale with the wavelength of the heating pulse? You used only a 1.06 micron laser to heat the sample. What if you varied that wavelength, how would you expect the results to vary? If the heat free electron heating theory is correct, then you would expect that the wavelength dependence would be very similar to the energy dependence, except that you replace e by the reciprocal of the frequency. Here is one of the immediate applications that you can make. Change the wavelengths, see if it still works. see the You have a problem with multiphoton generation, because in order to I would be very surprised if you saw multiphoton effects in Sio2 at a wavelength of half a micron, of the experiment to try inserting the second harmonic of YAG and see what one gets. Manuscript Received 1. Optical Measurements At PMS Electro-Optics Ramin Lalezari and Scott Knollenberg PMS Electro-Optics, Inc. Rapid and accurate measurements of scatter and absorption losses of optical elements provide essential feedback required for the development and production of low loss elements. Two techniques for the measurement of total loss and total integrated scatter are discussed. The total loss is determined using measurements of the resonant power in a two-mirror versus three-mirror active laser cavity. The difference in the power between a two- and three-mirror cavity can be directly related to the added third (test) element's loss. This simple apparatus will be described. The details of the measurement and some data on very low loss Ion Beam Sputtered mirrors will be presented. An apparatus developed for the measurement of total integrated scattered light from low-loss mirrors will be described. The instrument, which was designed to measure TIS levels of 1 to 10-5, consists of two coaxial parabolic mirrors with the sample located at the focal point of one mirror and a detector at the focal point of the second mirror. A chopped laser source and a filtered amplifier allow for the accurate measurement of a few ppm of scatter. Calibration is straight forward using a "white" diffuse scattering test specimen. Key Words: cavity loss; loss measurement; low-loss optics; mirrors; scatter measurement; total integrated scatter. Introduction Loss measurements on optical elements provide the feedback required for the development of critical optical elements and intercavity components specifically. Measurements of three of the following four quantities are helpful in characterization of the losses of a component (1-R for mirrors and 2-R for windows): total loss, absorption, scatter, and transmission (reflection for windows). Direct measurement of absorption is difficult and requires significant instrumentation. It is also greatly dependent on the thermal characteristics of the film and the substrate and its size and shape. Absorption is therefore only deduced from other loss measurements and not measured directly. Total loss, scatter, and transmission measurements can be made with relatively simple instrumentation and accuracy adequate for the testing of a large group of components, including low-loss mirrors for Ring Laser Gyroscopes. 2. Total Loss Measurement Total loss measurements can be made most accurately using resonant cavities. In 1984, Anderson et.al.described a mirror reflectometer based on optical cavity decay time. [1] This measurement technique is extremely sensitive to optical losses and is widely used to determine the losses of a variety of low-loss elements including RLG mirrors. This measurement requires relatively fast electronics (10 nsec response time) and great care in mode matching of the source laser beam to the test cavity. Another commonly used technique for measurement of losses of a mirror is to construct a scanning Fabry Perot interferometer from the mirrors and measure its finesse and throughput. This technique is also very sensitive to the mode matching of the source beam to the cavity and to the mechanical stability of the cavity. For extremely high finesse cases (10 ̄5), it becomes difficult to measure the finesse of throughput. The measure ment technique developed at PMS utilizes the power buildup in a saturated helium neon tube to determine the "Q" of the resonant cavity, which is inversely proportional to the cavity losses. This measurement is done at 632.8nm. The major assumption in the measurement is that the gain of the laser tube is constant over the range of measurement. The total cavity loss is measured for a two mirror high finesse cavity, including loss of the laser medium, Brewster's window, and capillary diffraction losses. The cavity is then folded by inserting the flat test mirror inside the cavity as indicated in the diagram (b). The drop in the resonant power is proportional to the increase in total losses which, is exactly twice the reflectance loss of the inserted mirror. The following steps are involved in the calibration and the measurement. 3. Reference Cavity Calibration Where P is measured power and K is the constant for the tube, K Po 0 1 + Ltube ; tube = 1 - TM11 (TM10 - TM12) Po The data is produced using the following reference cavity. See diagram (a). Test mirror M2 is inserted in the cavity as shown in diagram (b). 2° Note that for a given test set up, P, and Q, are constants and different elements can be 1 1 measured by aligning the cavity and measuring P. A similar calculation can be used to measure insertion loss of an AR coated window or Brewster's angle window from the reduction in Q of the reference cavity. 5. Error There is some predictable uncertainty due to the nonlinearity of the power produced in the tube as a function of cavity Q. This error can be understood by considering the following equation: [2] An estimated value of 0.04 for go and 0.0061 for L of the reference cavity will cause that cavity's losses to be 15% higher than the linear case where it is assumed that Pr = Ps 90/L. ± 1% uncertainty in the measurement of the power and laser noise due to mode sweeping creates a 2% uncertainty in the Q measurement and respectively a 2% uncertainty in the total cavity loss. There is also some error due to variations in gain with the mode structure conforming to defect patterns on the test element. This effect is generally not present except on elements with large defects. These errors can be avoided by monitoring the beam profile. A simple scatterometer is constructed for rapid measurement of scattered light from highly reflective dielectric mirrors. The scatter measurement is only approximate since the near an gle (0.2 steradians) scatter is not collected on the detector. Even though the instruments' measurements should not be compared to other instruments with different geometrical configuration, it has provided an extremely rapid and reproducible measurement for process development and quality control purposes. See diagram (c). The scatterometer utilizes two coaxial 4" diameter parabolic mirrors with a spider web fixture in the center which holds the sample and the detector in the focal points of the two A diaphragm and a flat black aperture provide a coarse but effective way of blocking the undesirable near angle scatter content of the illuminating beam. The laser beam is modulated with a 1KHz liquid crystal switch, and the detector amplifier has a 1KHz band pass filter. A lock-in amplifier would enhance the signal-to-noise ratio of the instrument. However, the improvements are not significant for an added cost of about 200%. The 1KHz signal from the detector is recorded on a digital oscilloscope and signal averaging is used in the case of low scatter mirrors, (less that 10ppm) to reduce noise. The system is calibrated using a M_0, coated element which is assumed to have 100% scatter. A five decade, 'g 2 two stage amplifier is used to provide necessary gain for the measurement. The following table presents some loss and scatter data from a variety of mirrors. |