ASTRONOMICAL SOCIETY OF THE PACIFIC 27 THE REFLECTING PROPERTIES OF ALUMINUM-. SURFACED MIRRORS. By Edison Pettit

Transcription

1 ASTRONOMICAL SOCIETY OF THE PACIFIC 27 THE REFLECTING PROPERTIES OF ALUMINUM-. SURFACED MIRRORS By Edison Pettit The recent development of a technique whereby a metal is deposited by evaporation upon a mirror surface in a high vacuum now makes it possible to prepare surfaces of some of the elements not heretofore available. The most promising metallic surface, both in general reflecting power and in permanency, is that of aluminum. The work of Dr. John Strong of the California Institute of Technology, described elsewhere in this number, has made it possible to prepare mirror surfaces of this metal up to three feet in diameter, whence it is probable that aluminum will become eventually a substitute for silver. With the expectation of using an aluminized siderostat mirror in a re-determination of the solar energy-curve in the ultraviolet, the reflecting power of this material has been investigated in a preliminary way on Mount Wilson. For the spectral region to the violet of À0.45ji, photographic methods were employed with an iron arc as source and a 1-meter concave grating spectrograph as the dispersion device. Timed exposures, variable sector, and variable source-distance gave sensibly the same results. 1 For wave-lengths to the red of À 0.45p, a radiometric method was employed. A Hilger constant-deviation monochromator and thermopile was fed with sunlight, a very satisfactory source on Mount Wilson. 2 The results, while not definitive, as they apply to only one pair of mirrors furnished by Dr. Strong, will be of interest to those planning to use evaporated aluminum surfaces. Table I shows the reflecting power of one aluminum surface at normal incidence, the reflecting power of a silvered mirror chemically deposited in the usual way, the sensitivity of the Eastman 40 photographic emulsion, the sensitivity of the eye, and the atmospheric transmission for Mount Wilson. The re- 1 Mt. W. ContrNo. 336; ApJ., 66, 43, Ibid.

2 28 PUBLICATIONS OF THE TABLE I Sensitivity. Atmos- Reflecting Power pheric Photographic Plate Trans- X Aluminum Silver (Eastman40)* -Eye mission 0.225m * Mees, J.O.S.A., 21, 767, December, fleeting power of the silvered mirrors was determined in the same manner as that of the aluminum and agrees closely with that obtained by others. The actual minimum reflectivity of silver is 0.04 at À 0.315p approximately. Figure 1 is a plot of the data in Table I with the addition of intermediate values available for some of the curves. The ordinates are decimal fractions, abscissae are wave-lengths in ui (10,000 angstroms). The light broken straight line between À 0.23p and À 0.6pt is the reflecting power of aluminum-surfaced

3 ASTRONOMICAL SOCIETY OF THE PACIFIC 29 mirrors as given by Williams and Sabine. 3 The reflecting power of chemically deposited silver is the light full line. The sensitivity curve of the eye is that for sources of high-surface brightness like that of a star; the sensitivity of the photographic plate (Eastman 40) is that given by Mees. 4 Fig. 1. (Above) The reflecting power of aluminum and silver. The broken straight line represents observations by Williams and Sabine. (Below) The transmission of the atmosphere above Mount Wilson, the sensitiveness of an Eastman 40 photographic plate, and of the eye. We may now compute the effects of a change from sil- ver to aluminum surfaces on the color index of a star and on the photographic and visual limits of the telescope. Since the silver and aluminum curves cross at À 0.4p, there will be no difference in color index or speed for this wave- length. For a star, however, the results will be determined by the integrated effects over the ranges of visual and photo- 3 Ap. 77, 316, J.O.S.A., 21, 767, 1931.

4 30 PUBLICATIONS OF THE graphic sensitiveness. These were obtained by multiplying the mean values, for intervals of 0.05//, of the transmission of the atmosphere, the reflecting power of the mirrors of the telescope, and the radiation from a black body at temperatures corresponding to various spectral classes. These products were then multiplied by the factors in the visual and the photographic sensitivity tables and the results summed up for each spectral class. The ratio of these sums expressed in magnitudes, with a constant subtracted to make the magnitudes zero at AO, will then be the color indices. Table II shows the color indices of stars of various spectral classes computed in this manner for the assumed temperatures TABLE II Type Temperature Absolute Centigrade Color Index (Mag.) 2 A1 Ag+Al A Color Index 2A1 (Ag+Al) Gain in Photographic Magnitude 2 A1 -(Ag+Al) Loss in Visual Magnitude 2 A1 (Ag+Al) BO... AO... P0... ggo.. gko.. gk * * Loss. given in the second column and for the various combinations of mirrors shown in the third, fourth, and fifth columns, where indicates two silvered mirrors ; 2 Al, two aluminum-coated mirrors ; and Ag -f- Al, one silver and one aluminum mirror in the telescope. The sixth and seventh columns show the corrections which must be applied to color indices determined with two aluminum mirrors or one silvered and one aluminumcoated mirror to reduce them to two silvered mirrors. These corrections are appreciable and must be taken into consideration when the mirror surfaces are changed. The last four columns show the change in photographic speed and visual sensitiveness when two silvered surfaces are

5 ASTRONOMICAL SOCIETY OF THE PACIFIC 31 exchanged for two aluminum or one aluminum and one silvered surface in the telescope. Photographically there is a gain in sensitiveness for all classes save gk5, which shows an insignifi- cant loss for silver plus aluminum. Visually there will be a loss for all spectral classes, amounting to only 0.08 mag. for late type stars when both mirrors are aluminum coated. The stars will appear bluer through a telescope provided with aluminum surfaces than through one with silvered mir- rors, as an examination of the curves in Figure 1 will show. The color indices in Table II, since they are greater for alumi- num than for silver, may at first sight lead to the opposite con- clusion. We can, however, understand the effect qualitatively in the following manner : Suppose filters are employed which transmit one very narrow band in the visual and one in the photographic region. Color indices determined with these fil- ters, if adjusted to be zero for type AO, would be in no way affected by reflections or absorptions in the instrument. If we imagine the two bands coincident, the color index will be zero, and if one band moves toward the violet, the color index will increase until with silver we reach the transmission band at À0.34p, when the color index will be a maximum. With alu- minum, however, the band can continue to the atmospheric cut- off at À0.30p. Hence for aluminum the color index should ex- ceed that obtained with silver. Carnegie Institution of Washington Mount Wilson Observatory January, 1934