INFORMATION TO USERS
This was produced from a copy of a document sent to us for microfilming. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the material submitted.
The following explanation of techniques is provided to help you understand markings or notations which may appear on this reproduction.
1. The sign or “target” for pages apparently lacking from the document photographed is “Missing Page(s)”. If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure you of complete continuity.
2. When an image on the film is obliterated with a round black mark it is an indication that the film inspector noticed either blurred copy because of movement during exposure, or duplicate copy. Unless we meant to delete copyrighted materials that should not have been filmed, you will find a good image of the page in the adjacent frame.
3. When a map, drawing or chart, etc., is part of the material being photo graphed the photographer has followed a definite method in “sectioning” the material. It is customary to begin filming at the upper left hand corner of a large sheet and to continue from left to right in equal sections with small overlaps. If necessary, sectioning is continued again—beginning below the first row and continuing on until complete.
4. For any illustrations that cannot be reproduced satisfactorily by xerography, photographic prints can be purchased at additional cost and tipped into your xerographic copy. Requests can be made to our Dissertations Customer Services Department.
5. Some pages in any document may have indistinct print. In all cases we have filmed the best available copy.
University Microfilms International
300 N. ZEEB ROAD, ANN ARBOR, Ml 48106 18 BEDFORD ROW, LONDON WC 1 R 4EJ, ENGLAND 8022334
R i n s l a n d , C u r t i s P h il ip
SPECTRA OF LATE-TYPE STARS IN THE ONE TO FOUR MICRON REGION
The Ohio State University Ph.D. 1980
University Microfilms
International 300 N. Zeeb Road, Ann Arbor, MI 48106 18 Bedford Row, London WC1R 4EJ, England PLEASE NOTE:
In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark
I, Glossy photographs
2c Colored illustrations
3, Photographs with dark background
'4, Illustrations are poor copy ___
5, Drin t shows through as there is text on both sides of page
6, Indistinct,, broken or small print on several pages ,^ throughout
7, Tightly bound copy with print lost in spii
8, Computer printout pages with indistinct print
9. Page(s) ______lacking when material received,, and not available from school or author ______
10. Page(s) ______seem to be missing in numbering only as text follows ______
11. Poor carbon copy ______
12. Not original copy, several pages with blurred type ______
13. Appendix pages are poor copy
14. Original copy with light type
15. Curling and wrinkled pages
16. Other
University MicroTiims international 30C \ 13 = 3 30 4=1 SQ 4 VI! 43'06 '313! 761-4700 SPECTRA OF LATE-TYPE STARS IN THE ONE TO FOUR MICRON REGION
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Curtis Philip Rinsland, B.S., M.S.
*****
The Ohio State University
1980
Reading Committee: Approved By
Professor Stanley J. Czyzak
Professor Geoffrey Keller
Professor Robert F. Wing
Department of Astronomy ACKNOWLEDGMENTS
My deepest gratitude goes to my adviser, Professor Robert F. Wing, for his encouragement and guidance during the course of this research project. His knowledge, insight, enthusiasm, and patience were indispensable in every phase of this work. If I am ever put in the role of advising graduate students, I only hope I can perform a small fraction as well as he has. His influence x^ill be appreciated long after this manuscript has gathered dust.
I am particularly grateful to Professor K. Narahari Rao for his assistance and encouragement during my many years as a graduate student.
The many beneficial ideas contributed throughout the pursuit of this work are greatly appreciated. His continued financial support has helped to sustain me during these lean times.
I am indebted to Dr. Sergio Ghersetti for making available the facilities of the Institute of Organic Chemistry in Venice, Italy. The hospitality extended to the author during his two visits to Venice is greatly appreciated.
Special thanks to Dr. Agostino Baldacci who introduced me to the mysteries of acetylene. It has been a pleasure to work with him both at The Ohio State University Infrared Molecular Spectroscopy Laboratory and in Venice, Italy. Without his advice and assistance, the analysis of acetylene reported in this work would not have been possible.
I would like to thank Drs. John H. Shaw and Robert J. Nordstrom of the Ohio State University for guidance to the literature of the atmospheric sciences.
I would also like to express my sincerest gratitude to the following people, places, and things for their contributions to this experiment and/or my life: Drs. S. Ridgway, D. Hall, and P. Connes
(for FTS data), Dr. C. Sneden (for MOOG), Kitt Peak National Observato ry (for the observing time), AUDREY (the rather cold and sometimes temperamental InSb detector of KPNO), Dr. R. Joyce (AUDREY'S father),
Dr. S. Giorgianni (whose wine cellar is one of the great landmarks of Italy), Dr. Z. Gastone (useful discussions), and Y. Gardner and
R. Bowman (great friends and neighbors).
Finally, I would like to thank my parents who have always encour aged me to do my best, offered understanding and support in all my decisions, and, above all, given me their love. VITA
December 5, 1950 . . . Born - Allentown, Pennsylvania
1972 ...... B.S. in Physics, The Pennsylvania State University, University Park, Pennsylvania
1972-1975...... Research Associate, The Ohio State University Radio Astronomy Observatory, Department of Electrical Engineering, The Ohio State University, Columbus, Ohio
1975 ...... M.Sc. in Astronomy, The Ohio State University, Columbus, Ohio
1975-1978...... Teaching Associate, Department of Astron omy, The Ohio State University, Columbus, Ohio
1978-197 9 ...... Teaching Associate, College of Mathemati cal and Physical Sciences, The Ohio State University, Columbus, Ohio
1979-198 0...... Research Associate, The Ohio State University Infrared Molecular Spectroscopy Laboratory, Department of Physics, The Ohio State University, Columbus, Ohio
PUBLICATIONS
"Ohio Survey VII", with R.S. Dixon, M.R. Gearhart, and J.D. Kraus, Astronomical Journal, 79, 1129 (1974).
"The Ohio Radio Sky Survey: Supplement 2", M.Sc. thesis, The Ohio State University, Columbus, Ohio (1975).
"Ohio Survey Supplement 2", with R.S. Dixon and J.D. Kraus, Astronomical Journal, 80, 759 (1975).
"Observations of First-Overtone Silicon Monoxide Bands in Cool Stars", with R.F. Wing and R.R. Joyce, in Symposium on Recent Results in Infrared Astrophysics, ed. P. Dyal, NASA TM-X-73, 190, 23 (1977). iv "Observations of First-Overtone Silicon Monoxide Bands in Cool Stars", with R.F. Wing and R.R. Joyce, in Symposium on Recent Results in Infrared Astrophysics, ed. P. Dyal, NASA TM-X-73, 190, 26 (1977).
"Scans of the 4-p SiO Bands in Late-Type Stars", with R.F. Wing, Bulletin American Astronomical Society, 10, 408 (1978).
"Atmospheric Extinction and Photometric Reductions near Four Microns", Bulletin American Astronomical Society, 10, 408 (1978)
"A Comprehensive Analysis of the Errors Affecting the Measurement of Intensities in Infrared Spectra", with D.W. Chen, K. Narahari Rao, W.C. Braun, and B. Fridovich, Thirty-third Symposium on Molecular Spectroscopy, Columbus, Ohio (1978).
"Strengths of 13C 150.? and 12C 1E,0?_ Lines at 4.3-4,5 pm", with W.C. Braun, B. Fridovich, G.R. Smith, E.E. Champion, D.W. Chen, and K. Narahari Rao, Thirty-third Symposium on Molecular Spectroscopy, Columbus, Ohio (1978).
"Aspects of Line Strengths and Half Widths for some the the V2 band lines of 12C 150 2 ", with V.M. Devi, P. Das, IC. Narahari Rao, and B. Fridovich, Thirty-fourth Symposium on Molecular Spectroscopy, Columbus, Ohio (1979).
"Acetylene Spectra Observed in N-type Stars: A Laboratory Study", with K. Narahari Rao, A. Baldacci, and S. Giorgianni, Thirty-fourth Sympo- ium on Molecular Spectroscopy, Columbus, Ohio (1979).
"Measurements of Atmospheric Extinction near 4 pm", with R.F. W.ng, Thirty-fourth Symposium on Molecular Spectroscopy, Columbus, Ohio (1979).
"Atmospheric Extinction in the 4 pm Region", with R.F. Wing, Astronomical Journal, 84, 1235 (1979).
"Strengths of 13C 1502 Lines at 4.3 pm", with A. Baldacci and K. Narahari Rao, Journal of Molecular Spectroscopy, in press.
v TABLE OF CONTENTS
Page ACKNOWLEDGMENTS ...... ii
VITA ...... iv
LIST OF T A B L E S ...... viii
LIST OF F I G U R E S ...... xiv
Chapter
I. INTRODUCTION ...... 1
1. General Comments ...... 1 2. Overview of the Program...... 9
II. SCANNER OBSERVING PROGRAM...... 14
1. Introduction...... 14 2. Instrumentation and Observing Procedure...... 17 3. Selection of Stars ...... 19 4. Data Reduction ...... 28
III. MEASUREMENTS OF ATMOSPHERIC EXTINCTION IN THE 4 ym REGION ...... 34
IV. OBSERVATIONS OF THE FIRST-OVERTONE SiO BANDS NEAR 4 ym . . 41
1. Introduction...... 41 2. Description of the Spectral Region ...... 43 3. Observations and Reduction Procedure ...... 49 4. Derived Quantities...... 52 5. Normal Oxygen-rich Giants and Supergiants...... 59 6 . Observations of Peculiar Supergiants ...... 65 7. Observations of Mira Variables...... 72 8 . Computation of Synthetic 4 ym Spectra...... 82 9. Comparison between Observed and Synthetic Spectra. . . 92
V. INTERPRETATION OF INFRARED COLORS...... 98
1. Introduction...... 98 2. Description of the Bandpasses...... 102
vi TABLE OF CONTENTS (CONTINUED)
Page 3. Observational D a t a ...... 107 4. Computation of Synthetic Colors...... 131 5. Comparison between Observed and Synthetic Color Temperatures for Giant and Bright Giant Stars...... 142
VI. LABORATORY STUDIES OF THE INFRARED SPECTRUM OF ACETYLENE . 158
1. Introduction...... 158 2. Theory of Vibration-Rotation Spectra of Linear Molecules...... 159 3. Measurements of Line Intensities of Linear Molecules . 165 4. Experimental Details ...... 176 5. Partition Function of 12C2H 2 ...... 185 6 . Analysis of the Spectrum at 3 p m ...... 188 7. Results of the 3 ym Intensity A n a l y s i s ...... 243 8 . Intensities of the Vi + V 51 and V 2 + V 51 Combination Bands and the Acetylene Column Density in IRC +10216 . 267
VII. CONCLUDING REMARKS ...... 275
APPENDIXES
A. Atlas of 4 ym Scans...... 278
B. Acetylene Line Positions in the 3 ym Region...... 294
LIST OF REFERENCES...... 315
vii LIST OF TABLES
Table Page
1. 17-point Scanner Program ...... 16
2. Program Stars...... 21
3. Vega Colors Computed with ATLAS...... 31
4. Comparison of Model Colors ...... 31
5. Summary of Observational Data...... 33
6 . Model Atmosphere Parameters...... 37
7. 4 ym Scanner Results for Non-Miras ...... 55
8 . Mean Relation between W(SiO) and MK Spectral Type for Giant and Bright Giant Stars ...... 61
9. Blackbody Indices in the 4 ym R e g i o n ...... 71
10. 4 ym Scanner Results for Miras ...... 78 o 11. Synthetic SiO Equivalent Widths (A)...... 93
12. Comparison of 2.1 ym Bandpass Depressions and 8-color CN I n d i c e s ...... 106
13. Narrow-band Magnitudes and Color Temperatures for Non-Miras...... 110
14. Error in Temperature Resulting from a 0.02 mag grror in the Color Index taken with respect to 10395 A . . . . 127
15. Effect of Reddening on Color Temperatures...... 129
16. Synthetic Colors and Blackbody Temperatures...... 133
17. Relations between Color, Color Temperature, Spectral Type, and Effective Temperature for Giant and Bright Giant Stars...... 147
viii LIST OF TABLES (CONTINUED)
Table Page
18. Classification of Vibrational States of Linear Molecules . 160
19. Infrared Selection Rules for Linear Molecules...... 164
20. Nuclear Spin Statistical Weights for Acetylene Isotopes of Point Group D ...... 164 “h 21. Experimental Details in Recording Acetylene Spectra. . . . 177
22. Experimental Conditions for 3 ym ^2^2 ^os-*-t:*-on Spectra • • 178
23. Ro-vibrational Constants (cm--*') and Statistical Weights for Lower Levels of 12(^2**2......
24. 12("2^2 Fartition Function...... 187
25. 12(^2^2 FartFtion Function Polynomial Coefficients...... 189
26. Calculated Unperturbed Energies for Several Diads and Triads of 12C2H 2 in Fermi Resonance (cm--*-)...... 193
-1 i ? 27. Molecular Constants (cm ) of i,iC„H Derived from Bands at 3 ym...... 194
28. Molecular Constants (cm ■*) of *-2C 13CH2 Derived from Bands at 3 ym...... 196
29. Observed and Calculated Wavenumbers (vac.cm--*) of the 0010°00-000000° Band of 12C 13CH2 ...... 198
30. Observed and Calculated Wavenumbers (vac.cm ^) of the 10011 0°e-0000°l1e Band of 12C2U2 ...... 199
31. Observed and Calculated Wavenumbers (vac.cm ^) of the 100110°f-0000°l1f Band of 12C2H 2 ...... 200
32. Observed and Calculated Wavenumbers (vac.cm ^) of the
0010000-000000° Band of 12C2H2 ...... 201
ix LIST OF TABLES (CONTINUED)
Table Page
33. Observed and Calculated Wavenumbers (vac.cm ) of the 001110°e-000110°e Band of 12C2H2 ...... 202
34. Observed and Calculated Wavenumbers (vac.cm of the 0011100f-000110°f Band of 12C2H2 ...... 203
35. Observed and Calculated Wavenumbers (vac.cm of the 0010°l1e-0000°l1e Band of 12C2H2 ...... 204
36. Observed and Calculated Wavenumbers (vac.cm of the 0010°l1f-0000°l1f Band of 12C2H2 ...... 205
37. Observed and Calculated Wavenumbers (vac.cm of the 010(ll)°+-000000° Band of 12C2H2 ...... 206
38. Observed and Calculated Wavenumbers (vac.cm of the 010(21)1e(TI)-000110°e Band of 12C2H2...... 207
39. Observed and Calculated Wavenumbers (vac.cm 2) of the 010(21)1f (II)-000110°f Band of 12C2H2 ...... 208
40. Observed and Calculated Wavenumbers (vac.cm 2) of the 010(12)1e(II)-0000°l1e Band of 12C2H2 ...... 209
41. Observed and Calculated Wavenumbers (vac.cm of the 010(12)1f(II)-0000°l1f Band of 12C2H9 ...... 210
42. Observed and Calculated Wavenumbers (vac.cm of the 0012°00-0002°00 Band of 12C2H2 ...... 211
43. Observed and Calculated Wavenumbers (vac.cm of the 001220°e-000220°e Band of 12C2H2 ...... 212
44. Observed and Calculated Wavenumbers (vac.cm of the 001220°f-000220°f Band of 12C2H9 ...... 213
X LIST OF TABLES (CONTINUED)
Table Page
45. Observed and Calculated Wavenumbers (vac.cm of the 001(11)°+-000(11)°+ Band of 12C2H2 ...... 214
46. Observed and Calculated Wavenumbers (vac.cm 2) of the 001(11)°_-000(11)°_ Band of 12C2H2 ...... 215
47. Observed and Calculated Wavenumbers (vac.cm of the 001(ll)2e-000(ll)2e Band of 12C2H? ...... 216
48. Observed and Calculated Wavenumbers (vac.cm 2) of the 001(ll)2f-000(ll)2f Band of 12C2H2 ...... 217
49. Observed and Calculated Wavenumbers (vac.cm 2) of the
0010°20-000002° Band of 12C2B2'’ ■•••••••••••• 218
50. Observed and Calculated Wavenumbers (vac.cm of the 0010°22e-0000°22e Band of 12C2H 2 ...... 219
51. Observed and Calculated Wavenumbers (vac.cm of the 0010°22 f-0000°22f Band of 12C9H2 ...... 220
52. Observed and Calculated Wavenumbers (vac.;cm of the 010(31)°+-0002°0° Band of 12C2H2 ...... 221
53. Observed and Calculated Wavenumbers (vac.cm of the 010(31)2e(II)-000220°e Band of 12C2H2 ...... 222
54. Observed and Calculated Wavenumbers (vac.cm of the 010(31)2f(II)-000220°f Band of 12C2H2 ...... 223
55. Observed and Calculated Wavenumbers (vac.cm 2‘) of the 010(22)°_-000(11)°- Band of 12C2H2 ...... 224
56. Observed and Calculated Wavenumbers (vac.cm of the 010(22)2e(II)-000(ll)2e Band of 12C2H2 ...... 225
xi LIST OF TABLES (CONTINUED)
Table Page
57. Observed and Calculated Wavenumbers (vac.cm of the 010(22)2f (II)-000(ll)2f Band of 12C?H2 ...... 226
58. Observed and Calculated Wavenumbers (vac.cm of the 010(13)°+-0000°2° Band of 12C2H2 ...... 227
59. Observed and Calculated Wavenumbers (vac.cm '^) of the 010(13)2e(II)-0000°22e Band of 12C2H2 ...... 228
60. Observed and Calculated Wavenumbers (vac.cm of the 010(13)2f (II)-0000°22f Band of 12C2H2 ...... 229
61. Observed and Calculated Wavenumbers (vac.cm 2) of the 0110000-010000° Band of 12C2H2 ...... 230
62. Observed and Calculated Wavenumbers (vac.cm of the 0011100e-000110°e Band of 12C 13CK2 ...... 231
63. Observed and Calculated Wavenumbers (vac.cm of the
001110°f-000110°f Band of 12C 13CH2 ...... 232
64. Observed and Calcualted Wavenumbers (vac.cm of the 0010°l1e-0000°l1e Band of 12C 13C1I2...... 233
65. Observed and Calculated Wavenumbers (vac.cm of the 00100l1f-0000°l1f Band of 12C 13CH2 ...... 234
66. Observed and Calculated Wavenumbers (vac.cm of the 010(ll)°+-0000°0° Band of 12C 13CH2 ...... 235
67. Observed and Calculated Wavenumbers (vac.cm 2) of the 1000000-000000° Band of 12C 13CHr ...... 236
68. Some Vibrational Term Values (cm of 237
xxi LIST OF TABLES (CONTINUED)
Table Page
69. Ground State Combination Differences for 12C 13CH2 . . . . 239
70. Observed Wavenumbers (vac.cm in the P Branch of the 100(11)2+-0000°2 ° Band of 12C2H2 ...... 244
-2 -1 71. Observed and Calculated Line Intensities (cm atm at 300 K) in the 0010°0°-0000000 Band of 12C2H2 ...... 246
-2 -I 72. Observed and Calculated Line Intensities (cm atm at 300 K) in the 010(11)2+-0000°0° Band of 12C2H2 ..... 247
-2 -1 73. Observed and Calculated Line Intensities (cm atm at 300 K) in the 001110°-0001100 Band of 12C2D-2 ...... 248
74. Observed and Calculated Line Intensities (cm —‘-atm 9 — ] ' at 300 K) in the 0010°11-0000°11 Band of 12C2H2 ...... 250
, -2 -1 75. Observed and Calculated Line Intensities (cm atm at 300 K) in the 010(21)1 (II)-000110° Band of 12C2H2 . . . . 252
76. Observed and Calculated Line Intensities (cm — 9 zatm — 1 at 300 K) in the 0 1 0 U 2 )1 (ID-OOOO0!1 Band of 12C2H2 . . . . 254
77. Observed and Calculated Line Intensities (cm — 9 ^atm — 1 at 300 K) in the 0010°00-0000°00 Band of 12C 13CH2 ...... 256
78. Summary of Acetylene Intensity Parameters Measured at 3 y m ...... 265
79. Comparison of 3 ym Integrated Intensity Values for Acetylene...... 266
-2 -1 80. Calculated Line Intensities (cm atm at 300 K) . . . . 270
xiii LIST OF FIGURES
Figure Page
1 . Comparison of observed and calculated extinction coefficients. Dots are the values measured at Kitt Peak on 1977 October 28 from observations of 12 standard stars. The smoothed curves show the calculated extinction coefficients for pressure- induced absorption by , far-wing absorption by CO2 , discrete line absorption by N90 , and continuous absorption by H 90. The level of the contribution by aerosol, wich is nearly constant throughout the spectral region, is indicated by the line segment in the lower right. The sum of the five components is also shown. This diagram has been taken from Wing and Rinsland ( 1 9 7 9 ) ...... 38
2 . Comparison of an observed 4 ym spectrum of atmospheric transmittance with calculated monochromatic trans- mittances at 10 cm-1 wavenumber intervals. The observed spectrum has been derived from an observation of the carbon star IRC+10216 on 21 October 1977 at Kitt Peak with a resolution of 0.05 cm 1 and a time averaged air mass of 1.085. The observed and calculated transmittances (solid circles) have been normalized so as to agree at 2500 cm-1 (4 ym). Con tinuous absorption by the molecules CO2 , and H2O has been included in the calculation of the synthetic transmittances also with aerosol extinction. The Vi + 2v2 band of N2O (center=2462.0 cm-1) is prominent in the observed spectrum...... 40
Kitt Peak Fourier transform spectrum of a Ori in the spectral interval 2442 to 2510 cm-1. The upper (reference) spectrum is of IRC+10216 which is feature less in this spectral interval except for telluric lines. The interference pattern present in the ref erence and stellar (middle) spectra is caused by a blocking filter. The lower spectrum is the ratio of the upper two and shows stellar features only. The locations of the R branch (2,0), (3,1), and (4,2) band- heads of 28SiO are labeled...... 46
xiv LIST OF FIGURES (CONTINUED)
Figure Page
4. Kitt Peak Fourier transform spectrum of a Ori in the spectral interval 2376 to 2445 cm-1. The format is the same as Fig. 3. The (5,3) and (6,4) o o bandheads of SiO are identified as well as the region affected by high J lines of the atmospheric V 3 band of CO2 . Beyond 2388 cm-1 the atmosphere is completely opaque...... 43
5. Smoothed 4 ym transformation coefficients ob tained on 28 Oct 77 (upper), 26 Oct 77 (middle) and 8 Nov 77 (lower) plotted vs. wavelength in air microns. The transformation coefficients are expressed on a magnitude scale per unit wavelength and have been smoothed using a Gaussian of half-width equal to the resolution (88 8 )...... 53
6. The relation between W(SiO) and MK spectral type for K and M stars. The peculiar supergiants discussed in Section 4.6 and Mira variables have been excluded. Luminosity class II and III stars are shown as solid circles, lab and lb supergiants as open circles, la and Ia-Iab supergiants as solid triangles, and MS stars as plus (+) symbols. The solid line represents the mean relation for luminosity class II and III stars...... 60
Scanner observations of 8 Gem (K0 Illb), ot Tau (K5 III), and 6 Oph (MO.5 III). The MK spectral types are indicated below the star name as well as the calculated I(104)-L(400) blackbody temperature. The wavelength scale is in air microns. The fluxes are on a magnitude scale per unit wavelength subject to arbitrary normalization. The bandpass (88 8 ), an 0.1 mag flux interval, and the positions of the OH and SiO features are shown below. Each scan is accompanied by a blackbody curve (solid line) of the temperature corresponding to the I(104)-L(400) color. The wavelength interval used to define the L(400) magnitude and to set the height of the blackbody curve is shown above the spectrum of a Tau...... 63 xv LIST OF FIGURES (CONTINUED)
Figure Page
8 . Scanner observations of (3 Peg (M2.5 II-III) , XY Lyr (M4-M5 II), and RX Boo (M8 :e) are shown on the same format as Fig. 7 ...... 64
9. Scanner observations of the M supergiants W Cep (M2 Iaep), 119 CE Tau (M2 Iab-Ib), and RW Cyg (M3-M4 Ia-Iab) are shown on the same format as Fig. 7...... 66
10. Scanner observations of the very luminous supergiants S Per (M4e la) and RW Cep (K0 O-Ia) are shown on the same format as Fig. 7. SiO emission may be present at the location of the SiO (2,0) and (3,1) bands...... 67
11. Scanner observations of the peculiar M supergiants NML Cyg, VX Sgr, and VY CMa are shown on the same format as Fig. 7 ...... 68
12. Scanner observations of the Mira variables R And (phase 0.77), o Get (phase 0.88), and IK Tau (phase 0.5) are shown on the same format as Fig. 7 ...... 73
13. Scanner observations of the Mira variables R Aur (phase 0.79), U Ori (phase 0.16), R Cnc (phase 0.87), and W Aql (phase 0.20) are shown on the same format as Fig. 7 ...... 74
14. Scanner observations of the Mira variables R Cyg and X Cyg are shown on the same format as Fig. 7. Upper observation was obtained at phase 0.34, middle at 0.31, and lower at 0 . 1 7 ...... 75
15. Scanner observations of the Mira variables x Cyg (phase 0.17) and R Cas (phase 0.68 and 0.63) are shown on the same format as Fig. 7...... 76
xv i LIST OF FIGURES (CONTINUED)
Figure Fage
16. A portion of a synthetic spectrum generated with MOOG. The model atmosphere used was on opacity-sampling log g = 1.0, T = 3200 K model. Solar elemental abundances, terrestria? isotope ratios, and a microturbulent velocity of 2 km/s were assumed in the calculations. The positions of atomic, SiO, OH and CN lines are indicated above and below. Molecular lines of the terrestrially most abundant isotope are indicated with arrows while short lines mark features from less abundant isotopes. The 20SiO bandhead at 2467.2 cm-1 is weakly visible in the computed spectrum...... gg
17. Continuation of the synthetic spectrum shown in the previous figure. The 28SiO (3,1) bandhead is at 2472.9 cm- 1 . The 29SiO bandhead at 2481.7 cm-1 is weakly visible in the computed spectrum ...... 37
18. Continuation of the synthetic spectrum shown in the previous figure. Most of the strong lines are due to the (2,0) band of 28SiO...... 88
19. Continuation of the synthetic spectrum shown in the previous figure. The (2,0) 28SiO bandhead occurs at 2497.2 cm-1. The region shortward of the (2,0) head is nearly free of line blanketing...... 89
20. Continuation of the synthetic spectrum shown in the previous figure. Most of the strong lines are due to OH. 90
21. Synthetic 4 pm scanner spectra are shown on the same format as Fig. 7. All spectra have been computed with solar elemental abundances, terrestrial silicon, oxygen, and carbon isotope ratios, and a 2 km/s microturbulent velocity. The effective temperature and surface gravities of the models are indicated to the left of the spectra. The uppermost model is from the grid of Bell _et _al. (1976a) . All other models are from Johnson, Bernat, and Krupp (1980) ...... 91
xvii LIST OF FIGURES (CONTINUED)
Figure Page
22. Plot of W(SiO) in Angstroms vs. effective temperature for solar-composition opacity-sampling models with different surface gravities and a fixed microturbulent velocity of 2 km/s. The positions of models with log g = 0 .0 , 1 .0 , and 2.0 are indicated with "x", "o", and "+" symbols, respectively. Sets of models with the same surface gravities have been connected with dashed lines. The solid line is the mean giant and bright giant relation taken from Table 8 . The effective temperature scale of Ridgway ejt al. (1980) has been used to convert from spectral type to effective temperature...... 95
23. Plot of W(SiO) in Angstroms vs. effective temperature for solar-composition opacity-sampling models of log g = 1.0 showing the effect of microturbulence. Open rectangles indicate the results obtained with a micro turbulent velocity of 2 km/s while solid rectangles represent values computed with 4 km/s. Sets of models with the same microturbulent velocity have been connected with dashed lines. The solid line is the mean giant and bright giant relation taken from Figure 22...... 96
24. Kitt Peak FTS spectrum of e Tau in the region of 2.101 pm bandpass of the infrared colors program. The top spectrum is a solar spectrum, the middle spectrum is the stellar spectrum, and the lower spectrum is the ratio of the two. Line positions are from the Hall (1970) a t l a s ...... 104
25. Comparison of an observed spectrum of a Ori (solid line) with a synthetic spectrum (+ symbols) in the region of the 2.101 ym bandpass of the infrared colors program. The region scanned was from approximately 4751 cm-1 (2.104 ym) to 4765 cm-1 (2.098 ym). Telluric absorption features are indcated below the spectrum. The synthetic spectrum was computed with CN lines only, ...... 105
xviii LIST OF FIGURES (CONTINUED)
Figure Page
26. Kitt Peak FTS spectrum of e Tau in the region of the 2.285 ym bandpass of the infrared colors program. The format is the same as in Fig. 24. The *2C0 (2,0) bandhead can be seen at 4360 cm- * ...... 108
27. Comparison between the continuous energy distribution of the Bell et_ _al. (1976a) solar composition T = 3750 K and log g = 2.25 model atmosphere (dotsf and a 3750 K blackbody energy distribution. The fluxes are on a magnitude scale per unit wavelength, normalized to the 1.0395 ym reference wavelength. Arrows mark the wavelengths of the infrared colors program and the continuum points of the Wing 8-color system. The peak in the continuum caused by the H- opacity minimum is clearly visible at 1.65 y m ...... 138
28. Absorption coefficient of H- (cm-* per neutral hydrogen atom) plotted against wavelength for a temperature of 3835 K and an electron pressure of 6.71x10“* dynes/cm2 . The solid line is the total H- absorption coefficient; the dashed line indicates the free-free component only. The conditions correspond to a level in the Bell et_ al. (1976a) Te = 3750 K and log g = 2.25 solar composition model close to continuum optical depth unity at 1.0395 y m ...... 139
29. The relation between T (104-210) and T (104-400). Stars of luminosity classes 5l and III are sfiown as + symbols and dots, respectively. Large open symbols represent models from the grid of Bell_et_al. (1976a), large solid symbols models computed by Johnson, Bernat, and Krupp (1980). All models were computed with solar abundances. Solid lines connect models with the same surface gravity. The dashed line is defined by the relation T (104-210) = T (104-400). The arrow indicates the approximate slope ofcthe reddening line...... 143
xix LIST OF FIGURES (CONTINUED)
Figure Page
30. The relation between T (104-400) and T (8c) for models and stars of luminosity class II and I?I, shown in the same format as Fig. 29. The dashed line is T (104-400) = T c(8c)...... 144
31. T (104-129) plotted against effective temperature. T^e solid line is defined by the luminosity class III relations in Table 17. The corresponding spectral types (+ symbols) are indicated. Opacity-sampling solar-composition model atmosphere results (Johnson, Bernat, and Krupp 1980) are indicated by solid symbols; open symbols mark values obtained with the solar-composition models of Bell _et _al. (1976a) . The dashed line defines the locus of points where the color temperature is equal to the effective temper- acure. . . . 148
32. T (104-210) vs . effective temperature is shown in the same format as Fig. 31. . 149
33. T (104-228) vs . effective temperature is shown in the same format as Fig. 31. . 150
34. T c (104-400) vs . effective temperature is shown in the same format as Fig. 31. . 151
35. Eight-color near-infrared color temperature vs. effective temperature is shown in the same format as Fig. 31...... 152
36. Correction for experimental loss. The singly cross- hatched region indicates the area under the line profile used to measure the equivalent width. Be cause of the residual line absorption near the ends of the integration interval, the 100% transmittance level will be underestimated. The additional area (doubly cross-hatched region) must be included to determine the true equivalent width...... 172
xx LIST OF FIGURES (CONTINUED)
Figure Page
37. Curves of growth. The data were taken from the tables of Jansson and Korb (1968)...... 175
38. Comparison of pressure gauge readings obtained with the NOAA/NESS and the Ohio State Baratrons. The data shown here were taken with the Ohio State 0-100 torr head and the NOAA/NESS 0-1 torr head. The percentage difference between the readings of the two gauges has been plotted vs. the pressure in pm Hg ...... 181
39. Profile of an absorption line obtained with the Ohio State University 10-m focal length Czerny-Turner vacuum grating spectrometer. The data were digitized with a Bendix Datagrid digitizer. Arrows mark the limits of integration used to determine the equiva lent width...... 184
40. Acetylene spectrum between 3290.1 and 3291.3 cm recorded at room temperature with a pressure of 3.05 torr and a path length of 96.95 cm. The stronger features are identified. The effect of nuclear spin statistics can be seen for rotational levels split by £-type doubling. The intensity ratio is 3:1 for lines of 12C2H 2 - ...... 190
41. Comparison between acetylene spectra recorded at 12.0 mm Hg pressure and 160 C (upper) and at 3.5 mm Hg pressure and room temperature (lower) in the region 3207 to 3211 cm- 1 ...... 192
42. Square of2the vibration-rotation matrix element I Rv 5 £' j ^ Pl°ttec^ against the integer m in the 0010°0°- 0000°0° band of 12C2H 2 . Error bars represent the standard deviation of the measure ments of an individual line. The horizontal . T a 1 t * I o dashed line is the mean value of Rv ~ 4 ...... 258 1 v £ J 1
xx i LIST OF FIGURES (CONTINUED)
Figure Page
43. Square of the vibration-rotation matrix element |RV n ^ 1^ vs< m in tbe 010(11)° -0000°0° band of 12C H . v £ J z z Error bars represent the standard deviation of the measurements of an individual line. The horizontal v » £»j? 2 dashed line is the mean value of |Rv | ...... 259
44. Square of the vibration-rotation matrix element IRv'£'J' I 2 VS’ m in the 010(21)1(II)-000ll°0 band of ^^C2^2' Error bars represent the standard deviation of the measurements for an individual line. The horizontal dashed line is the calculated value of ,v' £' J ’ 12 , ^ i^v' £ ,2 |R I obtained with the values of [Rr . [ and A "V A; J "V ~ listed in Table 78 ...... 260
45. Square of the vibration-rotation matrix element
\T? Z j '|2 vs. m in the 010(12)1 (II)-00000!1 band of 1 ? 2b2 ' Error bars represent the standard deviation of the measurements for an individual line. The horizontal dashed line is the calculated value of I _v ' £' J' 12 , . , . , , . £ ij't' 12 , . R „ , obtained with the values of R . and A 1 v £ J 1 ' v £ listed in Table 78 ...... 261
46. Square of the vibration-rotation matrix element |RV ln ? I2 vs. m in the 001110°-0001100 band of v £ J 1 C2b2 * Error bars represent the standard deviation of the measurements for an individual line. The horizontal dashed line is the calculated value of l^v'Ji'J’ ^ , . , .. , . |_v^ £ ^ 12 R . T obtained with the values of R . and A 1 v £ J 1 1 v £ 1 listed in Table 78 ...... 262
xxii LIST OF FIGURES (CONTINUED)
Figure Page
47. Square of the vibration-rotation matrix element i v ' f' T' 12 0 1 0 1 R I vs. m in the 0010 1 -0000 1 band of ' v £ J 1^C2H2 « Error bars represent the standard deviation of the measurements for an individual line. The horizontal dashed line is the calculated value of |RV ^ ^ obtained with the values of |RV „ and A 1 v £ J 1 1 v £ 1 listed in Table 78...... 263
48. Spectrum of the vj + V 5I band of 12C2H 2 near 4091 cm- 1 . The solid line is the laboratory spectrum measured with a pressure of 2.00 torr and a path length of 96.95 cm. The synthetic spectrum (+ symbols) has been con volved with a 0.030 cm-1 gaussian instrumental profile. Line positions and alternation of intensities are indicated below...... 271
49. Observed and synthetic spectra of the V 2 + V 51 band of 12C2H 2 near 2701 cm-1 are shown in the same format as Fig, 48. The laboratory spectrum was measured at a pressure of 5.00 torr and a path length of 96.95 cm. . . . 272
50. Scanner observations of a Lyr, a Ari, and S Oph are shown on the same format as Fig. 7 ...... 279
51. Scanner observations of a Boo, e Peg, and 6 And are shown on the same format as Fig. 7 ...... 280
52. Scanner observations of y Aql, p Per, and 8 Cnc are shown on the same format as Fig. 7 ...... 281
53. Scanner observations of £ Aur, a Lyn, andT 1 Aur are shown on the same format as Fig. 7 ...... 282
54. Scanner observations of u Gem, HR 1009, and y Eri are shown on the same format as Fig. 7 ...... 283
xxiii LIST OF FIGURES (CONTINUED)
Figure Page
55. Scanner observations of HR 8726, u Aur, and a Cet are shown on the same format as Fig. 7 ...... 284
56. Scanner observations of a Ori, tt Leo, and HR 1155 are shown on the same format as Fig. 7...... 285
57. Scanner observations of KQ Pup, 119 CE Tau, and y Cep are shorn on the same format as Fig. 7...... 286
58. Scanner observations of AD Per, p Per, and y Gem are shown on the same format as Fig. 7...... 287
59. Scanner observations of p Gem, tt Aur, and o 1 Ori are shown on the same format as Fig. 7...... 288
60. Scanner observations of SU Per, 51 BQ Gem, and HR 8621 are shown on the same format as Fig. 7...... 289
61. Scanner observations of p Per, XY Lyr, and RS Per are shown on the same format as Fig. 7...... 290
62. Scanner observations of 71 Peg, R Lyr, and HD 11961 are shown on the same format as Fig. 7...... 291
63. Scanner observations of a Her, EU Del, and 45 RZ Ari are shown on the same format as Fig. 7 ...... 292
64. Scanner observations of HD 207076 and HR 1105 are shown on the same format as Fig. 7...... 293
xx iv CHAPTER I
INTRODUCTION
Section 1.1 - General Comments
The multitude of phenomena present in the atmospheres of red giant and supergiant stars make them fascinating and particularly important sources of astrophysical information. If we adopt the most general definition of a stellar atmosphere as the region that connects the core of a star to the interstellar medium, red giant atmospheres reveal diverse processes such as variability, grain formation, shock waves, and mass loss which are of interest to theoreticians and observers alike.
In recent years the study of red giant stars has been advanced to a large extent by studies in the infrared spectral region. Although the observed IR spectra of many cool stars are complex, the infrared reveals important processes inaccessible by visual spectroscopy.
The purpose of this section is to discuss the difficulties connected with the interpretation of cool star spectra and to point out the role that IR studies can play in promoting an understanding of cool stars.
Many of the valuable contributions that have been made by IR investiga tors will be mentioned. An exhaustive review is not intended here since excellent review articles have appeared in the literature (Spinrad and
Wing, 1969; Merrill and Ridgway, 1979). Instead, since the region 1.5 to 4 ym has been investigated in this thesis , problems related to the interpretation of cool star spectra in this region will be emphasized. The formation of molecules at the low temperatures of the atmos
pheres of late-type stars has numerous effects on both observation and
interpretation. The dissociation equilibrium for most observable mole
cules is greatly affected by the carbon to oxygen abundance ratio. In
spectral appearance, stars with carbon-to-oxygen ratios less than one
are dramatically separated from stars with carbon-to-oxygen ratios
greater than unity. This effect is due to the very high dissociation
energy, 11.1 eV, of the CO molecule. If carbon is less abundant than
oxygen, all the carbon is bound in CO and the excess oxygen forms oxides
producing the "oxygen-rich" K and 15 star sequence. If carbon is more
abundant than oxygen, all the oxygen is tied up in CO and the left-over
carbon forms molecules such as CN, CH, and C2 , resulting in typical
carbon star spectra.
An important advantage of the presence of molecules is that isotope
shifts for both vibration-rotation and electronic transitions are much
larger than the shifts in atomic spectra. Frequently, these shifts are
sufficient to separate isotopic bandheads even in low resolution scans
(cf. Fay, et al. 1974). The 12C/13C ratio, which is a quantitative measure of the amount of nuclear-processed material reaching the surface
has been studied with several different approaches in the infrared.
High-resolution photoelectric scans of the red and near-infrared CN bands have been interpreted with curves of growth to derive 12C/13C
ratios in a large sample of G, K, and M type giant and supergiant stars
(Tomkin and Lambert, 1974; Tomkin, Lambert, and Luck, 1975; Dearborn,
Lambert, and Tomkin, 1975; Hinkle, Lambert, and Snell, 1976; Tomkin,
Luck, and Lambert, 1976). The observed values have been directly 3 compared to stellar evolution predictions. Although values of 12C/13C in the range 25 to 30 can be understood in terms of mixing as a star ascends the giant branch, problems remain in explaining some of the lower values observed (Dearborn, Eggleton, and Schramm, 1976; Scalo and
Miller, 1978). For fainter stars crude measurements of 12C/13C can be obtained rapidly with photoelectric scanner measurements of CN between
7820 and 9048 X (Wing, 1974).
The carbon monoxide vibration-rotation bands provide an important check on the 12C/13C ratios obtained from other molecules. Generally three sequences of CO bands are visible in the infrared spectra of cool stars. In order of decreasing strength, these are the fundamental
(Av=l) near 4.7 pm, the first-overtone (Av=2) near 2.3 pm, and the second-overtone (Av=3) near 1.6 pm. The greatest difficulty usually encountered in using the CO bands for isotope studies is the problem of severe line saturation. Carbon monoxide is fully associated in stars later than about K 3 , and very large CO column densities exist in many late type stars. The large line strengths observed in the first-over- tone region have lead to difficulty in extracting the 12C/13C ratio
(Thompson, 1973). The second overtone line oscillator strengths are two orders of magnitude weaker than the first overtone lines, and this sequence should be used when sufficient column density exists for the
13C0 lines to be observable. Spectral synthesis has been used in this region to derive the 12C/'3C ratio for a Ori (Lambert, Dearborn, and
Sneden, 1974; Gautier, jrt al., 1976) and a lower limit for 12C/13C in a Her (Thompson and Johnson, 1974) , although the results of the two a
Ori studies are contradictory. The discrepancy may be due to the assumed value of microturbulence (7 km/s by Lambert, Dearborn, and
Sneden, 1974; 2 km/s by Gautier, et_ a_l. , 1976), inadequacies in the model atmosphere structure, and/or errors in locating the continuum in the lower resolution (0.5 cm-1) spectrum of Gautier, et_ a\_. (1976). Use of curve of growth techniques has been found to give consistent results from the Av=2 and Av=3 bands of CO and CN lines near 8000 % for a Ori,
6 Peg, x Cyg, a Her, and a Boo (Hinkle, Lambert, and Snell, 1976).
Carbon isotope ratios derived for carbon stars are frequently con tradictory [compare, for example, the values of Fujita and Tsuji (1977) with the results of Climenhaga, et_ al.. (1977)], although it is encourag ing that Dominy, et a_l. (1978) obtained consistent values from CO and
CN. Other elements whose isotopes have been studied in the infrared include titanium (Lambert and Luck, 1977), silicon (Beer, Lambert, and
Sneden, 1974; Ridgway, Hall, and Carbon, 1977), and oxygen (cf. Maillard,
1974).
The quantitative measurement of molecular band strengths by means of a scanner or narrow-band filters has proved to be a very effective method of spectral classification. The most widely used and best cali brated system in use today is the 8-color system of Wing (1971). This system measures five quantities in the 0.7 to 1.1 qm region: the 1(104) magnitude, the continuum color, and the strength of TiO, V0, and CN.
This system has been applied to the classification of M supergiants
(White, 1971; White, 1972; White and Wing, 1978), carbon stars (Baumert,
1972) and M dwarfs (Wing, 1973). A modified 8-color system which measures ZrO, TiO, LaO, CN, 1(104), and continuum color has been devel oped by Piccirillo (1977) to classify and interpret S star spectra. 5
At longer wavelengths, quantitative classification is not as well
developed. Classification at wavelengths longer than 1.1 pm has been
discussed in two recent articles (Wing, 1979; Merrill and Ridgway, 1979).
Narrow-band photometry can also be used to derive atmospheric
abundances. Hartoog, Persson, and Aaronson (1977) have used intermediate
bandwidth filters to measure CO band strengths in weak G-band stars.
Their results confirm that these stars are deficient in carbon by a
large factor. The system of Piccirillo (1977), besides classifying
S-stars, has been calibrated with model atmospheres to derive 0/C abun
dance ratios.
The existence of strong blanketing by both atomic and molecular
lines creates severe difficulties in the computation of model atmo
spheres. In order to account properly for the bound-bound opacity,
literally millions of lines must be included in the. computations. The major opacity sources must be identified, accurate line lists with the
correct oscillator strengths must be assembled, and the opacity must be
accounted for properly and in a manner suitable for calculation in a
reasonable amount of computer time.
Considerable progress has been made in treating molecular line
opacities. Early model atmospheres (cf. Tsuji, 1967, 1969; Auman, 1969)
represented the opacity within a wavelength interval by a single number,
either a straight mean or a harmonic mean of the opacity. The straight mean introduces serious error in that the averaging process fills in the
low opacity windows between strong spectral lines where mujh of the
stellar flux emerges in a real stellar atmosphere. The harmonic mean,
computed from the average of the reciprocal opacities over a spectral 6 interval, is a better representation but the opacity is systematically underestimated and serious errors may result near the surface (Carbon,
1974).
Two methods are in current use which represent the opacity much more accurately. The opacity distribution function (ODF) method uses a series of numbers to approximate the distribution of the opacity within a wavelength interval. The advantages and disadvantages of this technique have been thoroughly discussed (Carbon, 1974; Van Paradijs and Vardya, 1975). An alternative and particularly flexible method called opacity sampling (OS) statistically samples the opacity at a set of points. Given a large enough grid of points, the computed model structure will approach the correct value (Peytremann, 1974; Johnson and
Krupp, 1976; Sneden, Johnson, and Krupp, 1976).
Current line lists used in the computation of model atmospheres include several million lines. Despite the large number, it is an impor tant observational task to determine if the computed models can correct ly reproduce the observed energy distributions. Accurate comparisons are required at both low and high resolution with spectroscopic and photometric techniques. Such studies should reveal if any important opacity sources have been omitted from the models. For many molecules, accurate laboratory data is needed for both model computations and spectral synthesis studies.
The infrared contains many important indicators of the CNO abun dance of late-type stars. Most of the molecules observed contain carbon, nitrogen, or oxygen; and the relatively low density of lines compared to the visible region in many stars makes the infrared potentially very useful. In oxygen-rich stars the molecules CO, CN, H 2 , OH, NH, SiO, HF,
and H20 have been identified while HCN, C2H 2 , CH, CN, CO, CS, and C2
have been detected in carbon stars. Many of the molecules seen in the
IR cannot be observed in the visible (a.g. CO, SiO), and hence the infra
red can provide unique information useful for CNO abundance studies.
In order to compute a synthetic spectrum to match a given star, the
effective temperature and surface gravity of the star must be accurately
known. For the K and M field giants, two recent studies have suggested
considerable revision of the effective temperature scale. Ridgway, et al.
(1980) have used angular diameters from lunar occultations and measured bolometric magnitudes to derive a temperature scale considerably warmer
than the one given by Johnson (1966), which has been in general use. The
revised scale is consistent with the results of Tsuji (1978a) based on a
comparison of observed and synthetic energy distributions. This agree ment suggests that (at least for the oxygen-rich giants) the effective
temperature scale may now be considered established. Although the sur
face gravities of the warmer late-type stars are sufficiently well known
to permit detailed spectrochemical studies (cf. Clegg, 1977; Lambert and
Ries, 1977), fundamental determinations of log g become extremely diffi
cult for the coolest stars owing to their vanishingly small parallaxes
and the absence of suitable binary star systems to derive stellar masses.
The astrometric measurements planned with the Hipparchus satellite and
the Space Telescope should improve the situation.
Measurements of energy distributions of celestial objects are
derived from comparisons with standard stars whose energy distributions
are known. The standard stars must be distributed throughout the sky and their flux distributions expressed on a uniform scale of absolute fluxes. Furthermore, the monochromatic flux differences between the standard stars must be accurately known if one is to use them to derive magnitudes or evaluate the stability of one's equipment.
No such data exist at wavelengths longer than 1.1 pm. The magni tudes which are available are essentially all wide-band measurements which include entire atmospheric windows. These data are of limited use for spectrophotometric reduction and in general do not have the photometric accuracy of fluxes at shorter wavelengths. The absolute fluxes of the wide-band magnitudes are not well established, and the extinction coefficients that have been measured with wide filters cannot provide the detailed knowledge of atmospheric extinction that is required for spectrophotometric calibrations.
Many additional difficulties exist in the interpretation of cool star spectra. Stratification effects, deviations from plane-parallel geometry and local thermodynamic equilibrium, micro- and macroturbulence, circumstellar shells, mass loss, and variability are just a few of the problems not completely understood. Despite what remains, considerable progress has been made recently and will no doubt continue to be made in the future. It is hoped that the results presented in this disserta tion will add to our understanding of cool stars. Section 1.2 - Overview of the Program
The purpose of this section is to discuss the objectives of this study and to describe briefly both the observing program and the methods used to interpret the data.
This investigation is primarily directed towards a study of the normal "oxygen-rich'' giant and supergiant stars of spectral types G, K, and M in the spectral region 1.2 to 4.1 pm. The infrared was selected because it is close to the energy maximum of these stars, is relatively free of line blanketing, and contains molecular bands of several of the most abundant molecules in stellar atmospheres, including CO, H 20, OH, and SiO. Meaurements of these bands are particularly useful for the determination of the elemental abundances of H, C, N, and 0, and their isotopes, and they may be helpful in the classification of extremely cool or heavily reddened stars not detectable at shorter wavelengths.
The objectives of this investigation are the following: (1) to mea sure the strengths of infrared bands of molecules containing carbon, nitrogen, or oxygen in a large sample of late-type stars to determine their suitability for classification purposes; (2 ) to study the energy distributions of late-type giants and supergiants by measuring narrow band magnitudes at a series of widely spaced continuum points throughout the infrared; (3) to compare the observed band strengths and continuum colors with the predictions of model atmospheres; and (4) to investigate the accuracy to which narrow-band photometry with a scanner can be done in the IR by making repeated measurements of a series of standard stars. 1 0
These objectives required the use of both low and high resolution
data. The Kitt Peak InSb grating spectrometer AUDREY has been used to measure molecular band strengths1 and narrow-band continuum magnitudes
in a large sample of late-type stars. This instrument offers flexi bility in the selection of wavelengths and bandwidths to be measured, moderate resolution (AA/A about 500), high speed, and good photometric accuracy. Observations were obtained during observing runs in Septem- ber-October 1976 and October-November 1977 with the Kitt Peak 1.3-m telescope.
The selection of program wavelengths requires a fairly detailed knowledge of the spectra of a series of representative stars. High- resolution Fourier transform spectrometer (FTS) spectra obtained with
the Kitt Peak 4-m telescope (Ridgwav and Hall 1978) and the M t . Palo- mar 5-m telescope (Connes and Michel 1974) were used, in combination with published observations, to select the program bandpasses. In addition, the FTS data have been used to study features requiring higher resolution than could be obtained with the scanner.
The observations have been compared to synthetic fluxes computed with model atmospheres and the synthetic spectrum program M00G (Sneden,
1974). The recently published grids of Bell, et_ aJ. (1976a) and John son, Bernat, and Krupp (1980) have been adopted and are believed to be the most realistic models currently available.
MOOG has been described in detail in the Ph.D. dissertation of
Sneden (1974) so that the procedure used to calculate the synthetic spectra will only be discussed briefly. The basic assumptions of MOOG are local thermodynamic equilibrium, plane-parallel geometry, and the 1 1
formation of lines in pure absorption. The model atmosphere parameters
(optical depth, temperature, gas pressure, and electron density) and
microturbulent velocity must be provided as input to MOOG along with
the relative abundances of the elements. With these parameters the
continuous opacity and molecular dissociation equilibrium can be calcu
lated at each layer in the atmosphere. A line list specifying wave
lengths, excitation energies, and gf values is required to evaluate the
line opacity at each depth in the atmosphere and at each wavelength in
the spectrum. From the continuum and line-plus-continuum optical depth
scale, MOOG computes the spectrum depths by integrating the second Milne
equation assuming the source function is the Planck function.
In addition to the studies of stellar spectra and energy distribu
tions, results of a laboratory analysis of the molecule C2H 2 are reported
in this dissertation. Acetylene has recently been detected in the
spectra of cool carbon stars in the 3 pm region (Ridgway, Carbon, and
Hall, 1978) and in the circumstellar spectrum of IRC +10216 (Ridgway,
_et _al. , 1976). The 3 pm depression is very strong in all cool carbon
stars and is due almost entirely to HCN and/or C2H2 . It is likely that
acetylene is an important opacity source in the upper atmosphere of many
of these stars and should therefore be included in model atmosphere cal
culations. To obtain the basic laboratory data required, the 3 pm bands
have been recorded with The Ohio State University 10-m focal length
Czerny-Turner vacuum grating spectrometer. Line identifications and
molecular constants are reported as well as measurements of the absolute
strengths of 101 lines in the seven strongest bands. The integrated
strengths of the + V 5 1 and v 2 + V 5 1 bands have also been measured and 1 2 have been used to derive a revised column density of acetylene molecules in the circumstellar shell of IRC +10216.
The results of this dissertation are presented in the following sequence. In Chapter II the scanner observing program is described in cluding the observing procedure and methods used to reduce the data.
In Chapter III are presented results of a study of atmospheric extinc tion in the 4 pm region. The extinction coefficients are the first narrow-band measurements reported in this spectral region. The observed extinction coefficients are compared to synthetic values generated with laboratory data and a model terrestrial atmosphere. Although the laboratory data were already available in the literature, they had not previously been assembled to calculate the effect of extinction on astronomical measurements in the 4 pm region and to determine the physical processes giving rise to this extinction. In Chapter IV scanner measurements of the 4 pm SiO bands are reported for 77 stars and are used to study the behavior of SiO strength with respect to spectral type and luminosity class. The SiO strengths observed in luminosity class II and III stars are compared to calculated values obtained with recently published model atmospheres. Previous studies of the SiO bands were limited to a much smaller number of stars. In
Chapter V the first narrow-band continuum color temperatures measured longward of 1 pm are reported and compared to model atmosphere pre dictions. The results of the high-resolution laboratory study of acetylene are presented in Chapter VI. The analysis at 3 pm is considerably more extensive and at higher resolution than previous studies and the observed line positions and absolute intensities 13 should be useful for quantitative studies of these bands in cool carbon star spectra. In Chapter VII the major results of this work are summarized, and a few concluding remarks are presented. CHAPTER II
SCANNER OBSERVING PROGRAM
Section 2.1 - Introduction
The Kitt Peak InSb grating spectrometer AUDREY has been used to study a sample of G, K, and M giant and supergiant stars in the 1.2 to
4.1 ym region. The observations have been reduced to yield measurements of the equivalent widths of the SiO first-overtone bands and narrow-band magnitudes at four widely spaced wavelengths nearly free of stellar blanketing.
The usefulness of narrow-band scanner measurements of cool stars in the near infrared has been amply demonstrated in the dissertation of
Wing (1967a). He used measurements at 27 carefully selected wavelengths between 0.7 and 1.1 ym to derive an infrared continuum magnitude (1(104)), a near-infrared color temperature, and band strength indices of the molecules TiO, VO, ZrO, and CN for over 300 stars. This system was later modified to be used with narrow bandpass filters by reducing the number of wavelengths to eight. The 8-color system (Wing, 1971) has been particularly useful for classifying stars of spectral types K, M, and C. Other applications have been discussed by Wing (1974).
Since narrow-band photometry of cool stars has proved to be so useful in the near infrared, measurements at longer wavelengths with the same procedure should also provide important information. The infrared beyond 1 ym is broken up by regions of very strong telluric
14 15
absorption, mostly by H 2O. Within the atmospheric windows, numerous bands of molecules containing H, C, N, and 0, and their isotopes, are observable. Measurements of these bands are of interest for CNO abund ance studies and may also be useful for classification in the IR. In addition, since there are spectral intervals which are nearly free of stellar line blocking in even very cool stars, narrow-band continuum magnitudes can be measured.
The original objective of these observations was to study the behavior of the molecular bands of CO, CN, SiO, OH, H2O, C2 , and SiO by making integrations at a relatively small number of wavelengths.
A 17-point photometric system (Rinsland, Wing, and Joyce, 1977) designed for this purpose was developed for the Kitt Peak scanner.
Unfortunately, flexure problems, described in Section 2.2, caused wave length shifts too large to be tolerated, and this program could not be carried out. Since any future system developed to study these molecules at a similar resolution will have to measure many of the same features, the wavelengths and bandpasses of the 17-point system are given in
Table 1. Measurements of these bands should prove useful for both classification work and for studies of CNO abundances.
Although measurements of spectral features cannot be made reliably with a single wavelength when the wavelength scale is uncertain, useful indices can still be derived from continuous scans over spectral fea tures. This procedure is less efficient but was necessary under the circumstances. Continuous scans have been obtained of the SiO first- overtone region, the first-overtone CO bands including the (2,0) 1 3C0 TABLE 1
17-POINT SCANNER PROGRAM
Central Central Wavelength Bandwidth Primary Wavenumber (cm-^-) in Air (ym) (cm"l) Order Feature* Contaminants
2493.7 4.0090 5.5 1 SiO (2,0) Atm. N2O 2496.0 4.0053 5.5 1 SiO (2,0) head Atm. N 2O 2502.0 3.9957 5.5 1 Continuum Atm. N 2O 4232.0 2.3623 7.9 2 ^ C O (4 >2) l3C0, Atm. H20 4263.0 2.3451 8.0 2 }3C0 (2 ,0 ) 12 C0, Atm. H 20 4354.0 2.2962 8.4 2 12 C0 (2,0) Atm. H 2O 4386.0 2.2794 8.5 2 Continuum -- 4518.5 2.2125 9.0 2 Continuum -- 4757.0 2.1016 10.0 2 Continuum -- 5647.0 1.7704 14.0 2 c2 Atm. H 2O 5713.0 1.7499 14.4 2 Continuum Atm. HoO, CO 5956.0 1.6785 10.4 3 Continuum CO 6253.7 1.5986 11.4 3 12C0 (5,2) Atm. CO2 6273.5 1.5936 11.5 3 Continuum Atomic lines, CO 6470.0 1.5452 12.4 3 Continuum CN 6501.0 1.5378 12.4 3 OH (4,2) CN 6512.0 1.5352 12.4 3 CN --
* In some bandpasses H2O will dominate in stars later than about M6 . 17 bandhead, the (5,2) CO bandhead, and the OH (A,2) R branch head. In this dissertation results of the SiO scans are reported.
For many program variables, concurrent observations on the 8-color system were obtained by R.F. Wing on a 0.4-m telescope at Kitt Peak.
These data were used to derive 1(104) magnitudes and spectral types.
Section 2.2 - Instrumentation and Observing Procedure
The infrared scanner observations were made at Kitt Peak National
Observatory with the 1.3-m telescope and a grating spectrometer (AUDREY).
The InSb photovoltaic detector was operated in the recommended zero bias setting and the liquid-nitrogen filled dewar was pumped to about 30 torr pressure to cool the system to 63 K to reduce detector Johnson noise.
Background subtraction was accomplished by square-wave modulation of the secondary mirror. For all observations the chopping frequency was
20 Hz and the chopping distance was usually set at 30 arcsec. Beam switching was used to eliminate, to first order, zero point drifts.
The entrance aperture used was quite large (18 arcsec) so that the photometric accuracy would not be limited by seeing fluctuations.
Scanning was accomplished by driving the grating with a stepping motor. The 0.5 mm exit slot employed for all measurements produced a band pass of about 88 R in first order. Observations were made in first order between 3 and 4 pm, in second order in the 2 pm region, and in third order at wavelength points shortward of 1.5 pm. A series of filters, cooled to liquid-nitrogen temperature, were used to block light from unwanted orders. For the brighter stars a series of two 18
or three integrations were taken at each wavelength with typical inte gration times of 10 seconds.
The operation of the scanner was computer-controlled. Observations were made in either a discrete wavelength or a continuous scan mode.
Continuous scans required as input a starting step position, a final step position, the separation in steps between each wavelength point,
and the number of integrations at each point. In the discrete wave length mode, the scanner performed integrations at a specified set of wavelengths, in both forward and reverse directions. Stepping motor position was calculated from the wavelength, the order, the stellar radial velocity, and two grating constants which were derived from calibration lines.
To calibrate the wavelength scale, hydrogen lines of the Paschen and Brackett series were observed in A-type stars. Unfortunately, during the first observing run in September 1976 it was discovered
that flexure in the scanner would cause uncontrollable wavelength shifts amounting to 10-20 percent of the widths of the bandpasses.
Wavelength shifts could be minimized by working in limited regions of
the sky and carefully centering the star with respect to the entrance aperture, but it was not possible to eliminate the problem. Observa tions of spectral features were therefore restricted to continuous scans, but continuum points in clean spectral regions could be observed on the discrete point program.
Although the majority of the data was obtained at night, some observations were made in the daytime. Since most of the stars 19
observed are bright visually, it was usually possible to see the stars
through the telescope. Stars fainter than about V=4 could be seen only under exceptional seeing conditions and acquisition had to be made by offsetting from bright stars and peaking up on the infrared signal of
the program star. Data acquisition was considerably slower for the invisible stars since scans had to be interrupted frequently to check
the centering. Another problem encountered in daylight observing was poor seeing. Generally, the seeing would begin to deteriorate seriously by 10 a.m. and remain poor until late afternoon. During times of poor seeing light would be lost outside the entrance aperture and observing had to be terminated. Some days of. good seeing did occur, and the photometric quality of these data was as good as obtained at night.
Section 2.3 - Selection of Stars
The selection of stars was governed by several factors. First, it was decided that the primary emphasis of the observing program would be to study normal oxygen-rich giant and supergiant stars of spectral types
G, K, and M. The observing list contained a large number of these stars well distributed in both spectral type and luminosity class between
G2 and M8 . Second, the stars selected for observation had to be relatively bright. AUDREY is a rather high resolution infrared scanner
(88 1 in first order with an 0.5 mm exit slot), and only for stars brighter than K~ 2 could scans be made within a reasonable amount of 2 0 telescope time. Third, it was felt that program stars should have well determined MK spectral types so that the dependence of the derived quantities with respect to temperature and luminosity could be accurately studied. And fourth, preference was given to stars that have been measured on the Wing 8 -color system. This was considered to be important since the 8-color system would provide narrow-band near-infrared measure ments of both color and color temperature that could be combined with the infrared scanner data to study the energy distributions of stars over a large wavelength interval (0.7 to 4.0 pm).
Fortunately, it proved to be possible to assemble an extensive observing list of stars which meet these criteria. For variable stars, whenever possible, simultaneous 8-color observations were made with a
0.4-m telescope at Kitt Peak. However, since 8-color measurements could not be made for daytime objects, not all of the variables have concurrent 8-color data.
In addition to normal oxygen-rich stars, some peculiar supergiants,
Mira variables, and chemically abnormal stars have been observed.
Several "infrared" stars such as NML Cyg and VY CMa were measured.
Since the observing time available at Kitt Peak was limited, it was not possible to study the variation of the derived quantities with phase in the Miras. A few MS, S, SC, and carbon stars were included in the program.
Table 2 lists the stars which have been observed with the scanner, in order of right ascension. In the first five columns the following information is contained: (1) the HD or BD number; (2) the HR or BS 2 1
Table 2
Program Stars
Variable Program ) or BD HR=BS Name Name Type Spectrum Col 4ym 8-c Notes
1013 45 X Peg M2+ III S X
1967 90 R And M S4, 6 e XX P=409.Od
3627 165 6 And K3 III X XX
3712 168 a Cas cst ICO- Ilia X X
4128 188 6 Cet 100058 K1 III X
6860 337 3 And 100088 M0 Ilia XX X
9927 464 51 And K3 III X
11961 M5 III XX
12929 617 a Ari 100163 IC2 11 lab S s X
14270 AD Per SRc M2.5 lab X X h&x Per
14386 681 0 Cet o Cet M M5 .5e XX P=331.6d
14469 SU Per SRc M3-4 lab X h&x Per
14488 RS Per SRc M4.5 lab X h&x Per
14528 S Per SRc M4e la X X h&x Per
17506 834 0 Per K3- Ib-IIa XX
18191 867 45 Ari RZ Ari I? M 6- III: XX X
18884 911 a Cet Ml.5 III S s
19058 921 P Per p Per SRb M4 Ilb-IIIax XX
20797 1009 M0+ Ila X X
22049 1084 e Eri K2 V X
22649 1105 328 S5,3 XX X 2 2
Table 2
Program Stars
(cont1d.)
Variable Program HD or BD HR=BS Name Name Type Spectrum Col 4ym 8 -c Notes
23475 1155 343 M2+ Ilab XX X
NHL Tau IK Tau M M9 X P=470d
25025 1231 Y Eri M0 III XX
27371 1346 Y Tau 102439 K0- Illab X
28305 1409 £ Tau K0 III XX
29139 1457 a Tau K5 III S s X
30959 1556 0 1 Ori 448 M3S XX X
32068 1612 ? Aur t, Aur EA K5 II+B XX
GP Ori SR SC X
32736 1648 W Ori SRb C5,3 X
34019 1707 R Aur M M6.5e X P=458.4d
36389 1845 119 Tau CE Tau SRc M2 Iab-Ib X X
38944 2011 u Aur 100686 M1+ Ilia X
39801 2061 a Ori a Ori SRc Ml-2 Ia-Ib X
39816 2063 U Ori M M 6 . 5e X P=372.4d
40239 2091 7T Aur 700 M3 II X X
42475 2190 TV Gem M0-1 lab X XX
42543 2197 6 Gem BU Gem Ic? Ml-2 Ia-Iab XX X
42995 2216 n Gem r) Gem SRb M3 III X X 23
Table 2
Program Stars
(cont1d.)
Variable Program I or BD HR=BS Name Name Type Spectrum Col 4pm 8-c Notes
44478 2286 y Gem 740 M3 Illab XX
44537 2289 Aur V 1 Aur Ic? K5--MOIab-Ib X
48329 2473 £ Gem 100759 G8 lb X
55383 2717 51 Gem BQ Gem M4 III XX
58061 VY CMa M5 I X x
60414 2902 KQ Pup M2 Iabep+B XX
60522 2905 U Gem M0 III XX
62509 2990 e Gem 100892 K0 Illb S s
69243 3248 R Cnc M M 6 ,5e X P=361.7d
69267 3249 6 Cnc K4 III XX
76827 3576 P UMa M3 Illb X
78712 3639 RS Cnc SRc M 6<2 (S) X
80493 3705 a Lyn 101039 K7 Illab X X
81797 3748 a Hyd 101049 K3 II-III X
82308 3773 X Leo 101056 K5 III X
84441 3873 £ Leo G1 II X
85503 3905 y Leo K1 .5 III XX CN1, Ca Str.
86663 3950 TF Leo M2-- Illab XX
90432 4094 y Hyd K5 III X 24
Table 2
Program Stars
(cont* d.)
Variable Program HD or BD HR=BS Name Name Type Spectrum Col 4ym 8-c Notes
95689 4301 a UMa 101174 K0-- Ilia S
96833 4335 UMa K1 III X
97778 4362 72 Leo 101191 M3 III X
100029 4434 X Dra 101207 M0 III X
108849 BK Vir lb M7- X
112300 4910 6 Vir M3 III X High vel.?
113226 4932 e Vir 101352 G8 Illab X
119228 5154 83 UMa 101398 M2 Illab X
124897 5340 a Boo 101433 K2 nip s s
126327 RX Boo SRb M 8 XX
127665 5429 P Boo K3 in X
131873 5563 B UMi 101477 K4 in X
132813 5589 RR UMi SR? M5 in X
140573 5854 a Ser K2 in X CN1.5
146051 6056 6 Oph M0,,5 III s
148387 6132 n Dra G8 III X
148783 6146 30 Her g Her SRb M 6-- Ill X
148856 6148 B Her 101593 G8 III X
156014 6406 a 1 Her SRc M5 Ib-II X
156283 6418 IT Her 101640 K3 Ilab X 25
Table 2
Program Stars
(cont'd.)
Variable Program HD or BD HR=BS Name Name Type Spectrum Col 4ym 8-c Notes
154143 6337 M3- III X
159181 6536 8 Dra G2 Ib-Ila X
161096 6603 8 Oph K2 III X X
164058 6705 y Dra K5 III X
165674 VX Sgr M4e X
167006 6815 104 Her 3895 M3 III X
i701^7 f' n X/XXVt /UU1 a Lyr 101745 A0 V o o X
172380 7009 XY Lyr lb M4-5 II X XX
175588 7139 62 Lyr 8 2 Lyr 1 M4 II X X
175865 7157 13 Lyr R Lyr SRb M5 III X X X
W Aql M S3 s9e X P=490.2d
185456 R Cyg M S3,9e XX P=426.3d
186791 7525 Y Aql K3 II s s X
187796 7564 X Cyg X Cyg M S7,2e X XX P=406.8d
+39°4208 RW Cyg SRc M3-4 Ia-Iab X XX
196610 7886 EU Del SRb M 6 III L-l X X High vel.
197989 7949 £ Cyg K0- III X X
IRC+40448 M 6 X NML Cyg
206778 8308 e Peg 102124 K2 lb X X
206936 8316 y Cep y Cep SRc M2 la X X X 26
Table 2
Program Stars
(con'd.)
Variable HD or BD HR=BS Name Name Type Spectrum Col 4ym 8-c Notes
207076 5468 M7 III: XXX
208816 8383 VV Cep EA M2 Ia-Iab XXX
210745 8465 X, Cep 102155 K1.5 lb X
212466 RW Cep Ic ICO 0-Ia XX
213310 8572 5 Lac 102179 M0 Iab+B X X
214665 8621 102195 M4 III XXX
216946 8726 102221 M0- lb XX
217476 8752 GO la XX
217906 8775 B Peg B Peg lb M2.5 II-IIIx XX
221615 8940 71 Peg 5749 M5- Ilia XXX
224427 9064 ip Peg M3 III XX
224490 9066 R Cas M M7e X X 27
number; (3) the Bayer or Flamsteed name or an alternate designation;
(4) the variable star name; and (5) the type of variability. Numerical designations in column 4 have been taken from the Catalogue of Stars
Suspected of Variability (Kukarkin, et al. , 1951), while the variability type refers to the designations of the General Catalogue of Variable
Stars (Kukarkin, et al., 1969). The spectral types listed in column 6 are from several references. For the G, K, and M stars the revised types of Morgan and Keenan (1973) have been used in preference to those given elsewhere. The spectral types of Mira variables except IK Tau refer to their normal maxima as given in the catalogue of Keenan,
Garrison, and Deutsch (1974). The spectral type of IK Tau is the mean maximum value derived by Wing and Lockwood (1973) on the basis of eight years of photometric measurements in the 1 pm region. Other references for spectral types were Buscombe (1977), the Catalogue of Bright Stars
(Hoffleit, 1964), Keenan (1954, 1963), Wildey (1964), Humphreys (1970), and Humphreys, Strecker, and Ney (1972). The columns headed "Col",
"4 pm", and "8-c" refer to the infrared colors program, the 4 pm con tinuous scans, and simultaneous 8-color measurements which were made
for the program variables, respectively. An "x" in these three columns means that data were obtained on the program while an "S" means that the star has been used as a standard in the reduction procedure.
The final column contains notes including the periods for all the Mira variables in days. 28
Section 2.4 - Data Reduction
In the reductions, all observations of standard stars obtained during any continuous observing period, corrected if necessary for wavelength shifts, were used in a least squares solution for the extinc tion and transformation coefficients for that period. The extinction was assumed to increase linearly with air mass, and fluxes have been expressed on a magnitude scale per unit wavelength interval.
The reduction programs of Wing have been used with minor modifica tions to reduce the data. Since the reduction procedure has been described in detail elsewhere (Wing, 1967a) , it will only be outlined here.
If the bandwidth of each measurement is sufficiently narrow that its effective wavelength is the same for all stars, and if the atmos pheric extinction depends linearly on the air mass, the flux per unit wavelength of star s at wavelength A and time t can be expressed on a magnitude scale by
F(s,A,t) = -2.5 log10 c(s,A,t,n,z) - k(A,n)X(z) + T(A,n) (2.4-1) where c(s,A,t,n,z) is the average counts per unit time (or the average d.c. signal) observed on night n at zenith distance z corrected for background; lc(A,n) is the extinction coefficient in mag per air mass;
X(z) is the air mass of the star; and T(A,n) is the transformation coefficient on a magnitude scale. In the reduction procedure X(z) has been approximated by sec z. 29
The transformation coefficient, which includes an arbitrary con stant, should be a smooth function of wavelength since it is a product of several functions that change slowly with wavelength: the detector sensitivity, the reflectivity of the optics, the efficiency of the grating, the transmittance of the filter, and the dispersion of the grating. An exception to this statement can occur in regions of large atmospheric extinction where features saturate and hence do not vary linearly with air mass. In those regions, the coefficients k(A,n) and
T(A,n) are no longer independent of z, but Eq. 2.4-1 still serves as a good first approximation since the lower atmospheric transmittance appears in the T(A,n) term.
In order to find lc(A,n) and T(A,n) a linear least-squares solution of Eq. 2.4-1 is performed with the observed readings c(s,A,t,n,z) for each standard star of known flux F(s,A,t), observed at air mass X(z).
For nights of lower photometric quality or for nights for which too few standard stars were observed for a reliable determination of the extinc tion and transformation coefficients, minimum and maximum permissible values of k(A,n) were specified. The assumed values of k(A,n) in these cases were based on nights where the observed coefficients were parti cularly well determined.
The absolute fluxes for the bandpasses used in the infrared colors program were derived with the model atmosphere program ATLAS (Kurucz,
1970) and the model structure computed for Vega by Schild, Peterson, and
Oke (SPO, 1971). The emergent colors computed with the OSU version of
ATLAS agreed to about 1% in the infrared with the published values of 30
SPO. The calculated monochromatic flux differences between each wave- o length and 10395 A, expressed on a magnitude scale per unit wavelength, are tabulated in Table 3. For the reduction of the continuous SiO scans a somewhat different procedure, discussed in Section 4.3, was used because of uncertainty in the hydrogen line strengths which affect the spectrum of Vega in the 4 pm region.
Recently Kurucz (1979) and Dreiling and Bell (1980) have published revised model atmospheres for Vega. Although the adopted model para meters differ (1^=9400 K, log g=3.95 for Kurucz; 1^=9650 K, log g=4.00 for Dreiling and Bell) , the computed infrared colors are very similar
to SPO. Dreiling and Bell (1980) note that their Vega model is about
2% fainter than SPO between 1 pm and 2 pm and agrees to within 1% for
2 pm < X < 10 pm. Model atmosphere predictions of infrared monochromatic
flux difference in magnitudes between 1.04 and 4.00 pm are compared in
Table 4. The models are designated in column 1 by their effective
temperature and surface gravity. The current status of the Vega calibra
tion problem has been discussed by Lange and Wing (1979).
The fluxes of the standard stars were obtained by iteration. The procedure was begun by reducing the nights of highest photometric quality on which Vega had been observed using mean extinction coeffi cients. By averaging the computed standard star fluxes for several nights, a preliminary standard star deck was assembled. These fluxes were then used to reduce the observations anew, without forcing mean
extinction coefficients. After assigning weights to the standard star
fluxes based on the quality of the night, several iterations were Table 3
Vega Colors Computed with ATLAS
Wavelength (£)
12870 0 . 8 6 8
21010 2.705
22850 3.052
33965 5.344
Table 4
Comparison of Model Colors
Parameters Reference I (104)-L(400)
9000/4.0 1 5.322
9400/3.9 1 5.353
9400/3.95 1 5.353
9400/4.0 1 5.349
9500/3.9 1 5.359
9500/4.0 1 5.356
9650/4.05 2 5.344
9650/4.0 3 5.370
^Kurucz (1979) 2 Computed with ATLAS using SPO parameters 3 Dreiling and Bell (1980) 32
required to minimize the residuals. To make all the program fluxes internally self-consistent, the final standard star deck was used to reduce all of the observations.
The probable errors in the magnitudes for the infrared colors program and the 4 pm scanner observations were 0.016 and 0.015 mag, respectively. The accuracy of the colors obtained in both programs was 0.008 mag. These values are typical of the errors quoted in the literature for photoelectric work and indicate that the Kitt Peak InSb detector is as stable against sensitivity changes as a good photo multiplier. The observational data are summarized in Table 5. 33
Table 5
Summary of Observational Data
IR Colors 4 ym Scans
No. of Standard Stars 10 11
No. of Standard Star Observations 42 48
N o . of Nights Used 5 6
No. of Days Used 2 3
No. of Stars Observed 1 1 1 * 77
Probable Error of Magnitudes 0.016 mag 0.015 mag
Probable Error o.f Colors 0.008 mag 0.008 mag
*34 stars observed at 1.29, 2.10, 2.28 ym only; 27 stars observed at 4 ym only; 50 stars observed at all four wavelengths of the infrared colors program CHAPTER III
MEASUREMENTS OF ATMOSPHERIC EXTINCTION IN THE 4 ym REGION
The Kitt Peak grating spectrometer AUDREY has been used to record stellar spectra from 3.98 to 4.07 ym. Although the primary purpose of the program was to study the behavior of the SiO first-overtone vibra- tion-rotation bands in a sample of late-type stars (Chapter IV), atmospheric extinction was measured by making repeated observations of a series of non-variable, standard stars. The extinction at 4 ym was found to be larger than expected and strongly wavelength dependent.
The results of this study have been published (Wing and Rinsland, 1979) and will be summarized here.
The region studied is near the long-wavelength edge of the photo metric L band. Inspection of high-resolution solar spectra of this region revealed only weak telluric lines of N 2O so that atmospheric extinction was expected to be small. The observations, however, revealed that the extinction in this region was suprisingly large, increasing from about 0.05 mag per air mass at 3.98 ym to 0.25 mag per air mass at 4.07 ym. The weak N2O bands did not have the proper distribution or sufficient strength to account for the general increase in extinction to longer wavelengths observed in this region.
To account for the measured extinction the following sources of atmospheric opacity were considered: continuous absorption by the
34 35
molecules N 2 , CO2 , and H2O; discrete line absorption by N 2O; and aerosol extinction. Absorption by N 2 in this region is caused by pressure-induced transitions in the fundamental vibration-rotation band. The pressure in the terrestrial atmosphere is sufficient to induce a dipole moment, and measurable absorption occurs because of the high concentration of nitrogen in the earth's atmosphere. Continuous absorption by CO2 is from the wings of distant lines from the v3 fun damental and its associated "hot" bands near 4.3 pm. Although the region observed is far from these bands, the superposition of the extreme wings of these very intense lines produces a strongly wave length-dependent extinction. Weak continuous absorption by water vapor has been observed in this region in the laboratory (Burch, Gryvnak, and Pembrook, 1971; Watkins, et al., 1979), but its nature is unclear.
The same process as for CO2 is likely to occur, although absorption from water vapor dimers is suspected and may dominate in some spectral regions. Absorption and scattering by small particles (aerosol extinc tion) produces only a small contribution in the infrared.
The observed extinction coefficients have been compared to synthe tic extinction coefficients calculated from absorption coefficients measured in the laboratory and a model terrestrial atmosphere. A mid latitude Spring-Fall model (U.S. Standard Atmosphere Supplements, 1966) was adopted. Water vapor content as a function of altitude was approximated by scaling the mean annual mid-latitude mixing ratios of
Gutnick (1965) so as to agree with the observed water vapor content at
Kitt Peak at the time of the observations. The adopted model 36
parameters are tabulated in Table 6 . Absorption from N2 O was evaluated from the atmospheric lines appearing in the interferometric spectrum
(resolution=0.05 cm-1) of IRC +10216 (Ridgway and Hall, 1978)
Aerosol extinction was estimated from the extinction observed at Kitt
Peak in the 1 ym region (which is believed to be due almost entirely to aerosols) and theoretical calculations of the wavelength dependence of aerosol extinction (Elterman, 1964).
The comparison of observed and computed extinction coefficients is shown in Figure 1. The calculations indicate that the primary com ponent of extinction in the region observed is N 2 . The N2O lines, primarily from the + 2v2 band, also produce considerable absorption in the region. The C02 absorption is small in the region observed, but it rises rapidly at wavelengths longer than 4.10 ym and reaches equality with the N 2 absorption at 4.17 ym, just before the onset of
C02 line absorption from the v 3 fundamental band. The contributions of water vapor and aerosol are small throughout this region and nearly independent of wavelength. Considering the many sources of error in the laboratory data, the synthetic extinction coefficients are in good agreement with the observations.
Since the photometric measurements of extinction were limited to a rather narrow wavelength interval (3.98 to 4.07 ym), it is of interest to compare the calculated coefficients with observations made over a larger spectral region. Ridgway (1979) and his associates have used a high-resolution (0.05 cm-1) spectrum of IRC +10216 to obtain measure ments of atmospheric transmittance in the 4 ym region. The observation 37 TABLE 6
MODEL ATMOSPHERE PARAMETERS
ALTITUDE TEMPERATURE , PRESSURE H_0 MIXING RATIO (1cm) (K) (mb)
2.00 275.15 795.0 0.01160 2.25 273.53 770.6 0.01079 2.50 271.90 746.8 0.01002 2.75 270.28 723.7 0.00927 3.00 268.65 701.1 0.00856 3.25 267.03 679.1 0.00788 3.50 265.40 657.6 0.00723 3.75 263.78 636.8 0.00661 4.00 262.15 616.4 0.00602 4.25 260.53 596.6 0.00547 4.50 258.90 577.3 0.00494 4.75 257.28 558.5 0.00445 5.00 255.65 540.2 0.00399 5.25 254.03 522.4 0.00356 5.50 252.40 505.1 0.00316 5.75 250.78 488.2 0.00280 6.00 249.15 471.8 0.00246 6.25 247.53 455.9 0.00216 6.50 245.90 440.3 0.00189 6.75 244.28 425.3 0.00165 7.00 242.65 410.6 0.00144 7.25 241.03 396.4 0.00127 7.50 239.40 382.5 0.00 1 1 2 7.75 237.78 369.1 0.00 1 0 1 8.00 236.15 356.0 0.00093 8.25 234.53 343.3 0.0 8.75 231.28 319.0 0.0 9.25 228.03 296.2 0.0 9.75 224.78 274.6 0.0 10.25 221.53 254.4 0.0 11.50 216.65 209.2 0.0 14.00 216.65 141.0 0.0 16.00 216.65 102.9 0.0 17.00 216.65 87.87 0.0 18.50 216.65 69.36 0.0 20.00 216.65 54.75 0.0 21.50 218.15 43.25 0.0 23.00 219.65 34.22 0.0 24.00 220.65 29.30 0.0 26.50 223.15 19.94 0.0 28.50 225.15 14.70 0.0 31.00 227.65 10.08 0.0 35.00 237.05 5.589 0.0 39.00 248.25 3.182 0.0 47.00 270.65 1.109 0.0 MAGNITUDE PER AIR MASS 0.4 0.3 0.2 0.5 0.0 0.1 which is nearly constant throughout this spectral region, region, spectral this throughout constant nearly is which en ae fo igadRnln (1979). Rinsland and Wing from taken been CO by absorption by H^O. The level of the contribution by aerosol, aerosol, by contribution the of level The H^O. by absorption extinction calculated and observed of Comparison 1. Figure is indicated by the line segment in the lower right. The The right. lower the in segment line the by indicated is The stars. standard 12 of observations from 28 October 1977 sum of the five components is also shown. This diagram has diagram This shown. also is components five the of sum coefficients extinction calculated the show on curves Peak smoothed Kitt at measured values the are Dots coefficients. for pressure-induced absorption by N _ , far-wing absorption absorption , far-wing _ N by absorption pressure-induced for 2 , discrete line absorption by N^O, and continuous continuous and N^O, by absorption line , discrete 2500 xicin Coefficients Extinction bevd s Calculated vs. Observed .00
W AVELENGTH AVELENGTH W VNME ( 1) " m (c AVENUMBER W 2450
(^m) 4.10 TOTAL CO. 4.15 2400 Aer. 38 39
was made at Kitt Peak National Observatory on 21 October 1977, with a
Fourier transform spectrometer and the 4-m telescope. The time averaged
air mass was 1.085. IRC +10216 is a late-type carbon star heavily
obscured by a dust shell, and its 4 pm spectrum is continuous except
for telluric lines. Corrections have been applied for the variation
of the response function of the interferometer with wavelength by
observing a blackbody source of known temperature. The variation of
stellar flux with wavelength has been corrected for by dividing the
observed spectrum by a blackbody having the photometric color tempera
ture of IRC +10216. The resultant transmittance spectrum has arbitrary
normalization.
The Kitt Peak spectrum is compared with synthetic extinction cal
culations in Figure 2. The 100% transmittance level has been derived
by requiring the observed and synthetic transmittances agree at
2500 cm-1 (4.0 pm). Calculated monochromatic transmittances (solid
circles), shown at 10 cm-1 intervals, have been computed as previously
described except that N2O line absorption was not included. The
observed and calculated transmittances are in excellent agreement
throughout the interval. rmnn i teosre spectrum. theobserved in + Viprominent The extinction. aerosol with along themolecules by absorption ’ anc* transmittances Continuous C02 synthetic ^2* the of t>een ^as ^2® calculation the in (4 been ym).Included have cm-1 2500 (solid circles)at to agree as so transmittances normalized calculated and observed The airmass 1.085. averaged timeof a and cm-1 0.05 of resolution a with Peak Kittat 1977 October pcrmhs endrvd rma osraino tecro sa IC126o 21on IRC+10216 star carbon the ofobserved The observation an from derived been has spectrum intervals. wavenumber cm-1 10 at transmittances monochromatic calculated
Figure 2. Comparison of an observed 4 ym spectrum of atmospheric transmittance with with transmittance atmospheric of spectrum 4ym observed an of Comparison 2. Figure sraA2g»“9“«— 3
1 f ■ 1 m m CHAPTER IV
OBSERVATIONS OF THE FIRST-OVERTONE SiO BANDS NEAR 4 pm
Section 4.1 - Introduction
The infrared spectra of many late-type stars exhibit absorption bands due to the first-overtone vibration-rotation sequence of the free radical silicon monoxide (SiO). Since SiO is predicted to be an abundant molecule in cool oxygen-rich giants and supergiants (cf.
Johnson, Beebe, and Sneden, 1975) and the first-overtone bands are conveniently placed for measurement in the L-band window, scanner observations were made to study the behavior of SiO with respect to spectral type, luminosity class, and chemical composition.
The first-overtone bands of 2 8 Si160 were first observed by
Cudaback, Gaustand, and Knacke (1971) in a Ori (M2 lab). At their resolution of 2 cm-1 the longward degrading R branch bands are clearly visible starting with the (2,0) head at 2497.2 cm-1. Crude synthetic spectra were used to show the SiO absorption was photospheric in origin.
Observations of the region of the (3,1) bandhead of 2 8 Si160 were made by Wollman et. al. (1973) using a Fabry-Perot interferometer mounted at the coude focus of the Lick 120-inch. At a resolution of
0.6 cm-1, absorption was detected in 9 M giants and supergiants and in the mild S-type Mira x Cyg (S7,le-SlO,le). Negative detections were
41 42 reported for the cool carbon stars Y CVn and V Cyg, the prototype long period variable o Ceti (M5-M8), the semiregular variable W Hya (M8 e), two peculiar M supergiants (NML Cyg and VX Sgr), and a Boo (K2 IIIp).
The first high resolution observations (0.1 cm-1) of SiO were reported by Beer, Lambert, and Sneden (1974). Their Fourier transform spectrum of a Ori fully resolved the rotational structure and confirmed that SiO was the dominant absorber in the region. A few stellar OH lines from the P branches of the fundamental bands were also identified.
A model atmosphere (Johnson, 1974) and spectral synthesis were used to estimate the SiO abundance and upper limits to the silicon isotope ratios.
The dependence of SiO strength with phase has been studied in the spectra (resolution 0.5 cm-1) of several long period Mira variables
(Hinkle, Barnes, and Lambert, 1976). In the three stars observed — o
Ceti (M5-M8), R Leo (M7-M9e), and y Cyg (S7,le-S10,le) — the SiO bandheads are not detectable at maximum light. Near minimum the band- heads of 28SiO were observed in all three stars. The shape of the continuum was also noticed to change with phase, although this effect may be due to the wavelength dependent extinction noted in this spectral region (see Chapter III). The hydrogen line Ba which lies near the
(3,1) SiO bandhead was observed in emission in some of the spectra obtained near maximum light. The isotopic bandheads of 2^SiO and 30SiO were detected in y Cyg, and crude spectral synthesis indicated a near terrestrial ratio of the silicon isotopes. 43
A number of stars have been observed near 4 ym at high resolution with the Kitt Peak FTS (Ridgway, Hall, and Carbon, 1977). The 30SiO bandheads are faintly visible in some stars and indicate a 2 8 Si/30Si abundance ratio consistent with the terrestrial value of 30. SiO has been seen in emission in some Mira variables.
The SiO bands are relatively weak features. SiO absorption cannot be seen in the low resolution spectra of Merrill and Stein (1976) and
Noguchi et al. (1977). Merrill and Stein (1976) did note a sharp rise in flux to longer wavelengths in the spectra of S Per (M4e la) and
IRC+ 60370 (K4.5 la) at the location of the SiO bands. Their suggestion of SiO emission in some highly luminous stars is supported by the results obtained in this study.
Preliminary results of this study have been reported previously
(Wing, Rinsland, and Joyce, 1977; Rinsland and Wing, 1978). A discus sion of the SiO results was also included in a review paper on classi fication by photometric measurement of molecular bands (Wing, 1979).
Section 4.2 - Description of the Spectral Region
The Kitt Peak grating spectrometer has been used to scan the spectral interval 3.98 to 4.07 ym in a sample of late-type stars at a - 1 0 resolution of 5.5 cm 1 (88 A). Of primary interest was to study the first-overtone vibration-rotation sequence of the silicon monoxide free radical. Within this interval are located the R branch bandheads of the
(2,0) and (3,1) bands of the primary isotope 2 8 Si160 at 2497.2 cm-1
(4.003 ym) and 2472.9 cm-1 (4.043 ym), respectively. In addition, the 44
isotopic (2,0) heads of both 29 Si160 and 30Si16O occur in this region.
The (2,0) 2 9 Si160 bandhead is favorably located at 2481.7 cm-1 (4.028 ym) between the main isotope bandheads. The (2,0) 30Si15O head occurs at 2467.2 cm 1 (4.052 ym) and is difficult to separate from the 2 8 Si150
(3,1) bandhead at the resolution used. The terrestrial isotope abun dance ratios of 2 ®Si:23Si : 3^Si are 92.2:4.7:3.1.
Also of interest were the hydrogen lines in this spectral interval.
The Brackett a line occurs at 2467.8 cm-1 (4.051 ym), very close to the location of the (3,1) 2 8 Si150 bandhead. Absorption from this line can be expected in early type stars (Dreiling and Bell, 1980) and was ob served on this program. Paschen and Brackett lines have been seen in emission in Mira variables in the 1 to 2 ym region (cf. Johnson, et al.,
1968; Johnson and Mendez, 1970) and Ba emission has been reported in x
Cyg (Hinkle, Barnes, and Lambert, 1976). Detection of Brackett a emission is difficult with low-resolution scans because of the coinci dence of the (3,1) 2 0 Si16O band. The hydrogen line Humphreys 14 at
2487.0 cm-1 (4.020 ym) can also been seen in the infrared spectra of
Sirius and Vega calculated by Kurucz (1979).
The published spectrum of a Ori (Beer, Lambert, and Sneden, 1974) and Kitt Peak spectra of a Tau, y Gem, and a Ori (Ridgway and Hall, 1978) were used to examine the spectral region at high resolution. The Kitt
Peak spectrum of a Ori is reproduced in Figures 3 and 4. In all three stars SiO is the dominant absorber with a small number of OH fundamen tal band P branch lines occurring throughout the region. Shortward of the onset of SiO absorption only a few OH lines are observable, and the Figure 3. Kitt Peak Fourier transform spectrum of a Ori in the spectral interval 2442 to 2510 cm-^. The upper (reference) spectrum is of IRC+10216 which is featureless in this spectral interval except for telluric lines. The inter ference pattern present in the reference and stellar (middle) spectra is caused by a blocking filter. The lower spectrum is the ratio of the upper two and shows stellar features only. The location of the R branch (2,0), (3,1), and (4,2) bandheads of 28SiO are labeled.
45 tfflVELEKGTH (BJ r c
kavtfac3£R its? ) 46 Figure 4 Kitt Peak Fourier transform spectrum of a Ori in the spectral interval 2376 to 2445 cm-'*'. The format is the same as Fig. 3. The (5,3) and (6,4) bandheads of 28SiO are identified as well as the region affected by high J lines of the atmospheric v 3 band of CO2 . Beyond 2388 cm”*- the atmosphere is completely opaque.
47 KSWELEKCTH (f5>
KSVEKUXEER 5 48 49
scanner magnitude observed at 3.996 ym has been used as a continuum point in the reduction procedure and in the program of infrared colors
(Chapter V).
The telluric spectrum has been discussed in detail as part of a study of atmospheric extinction in the 4 ym region (Wing and Rinsland,
1979; Chapter III). A high resolution spectrum of atmospheric trans mittance near 4 ym is shown in Figure 2.
Section 4.3 - Observations and Reduction Procedure
Continuous scans between 3.98 and 4.07 ym have been obtained of o _. 77 late-type stars at a resolution of 88 A (5.5 cm 1). Measurements u were made at 51 spectral points separated by a step size of 16 A
(=0.0016 ym) or 1 cm- 1 , corresponding to about 5 data points per resolution element. Observations were made during observing runs in
September-October 1976,and in October-November 1977. During the latter run, near simultaneous 8-color observations were made with the //3 Kitt
Peak 0.4-m telescope of most of the brighter program variables accessi ble during nighttime. These measurements provided concurrent 1(104) magnitudes and 8-color spectral types.
The basic scanner data reduction procedure has been described in
Section 2.4. Flexure in the scanner caused the wavelength scale to shift from star to star by amounts up to several steps. However, the presence of SiO features in the data allowed the step positions to be re-numbered, and the uncertainty in the wavelength scale was reduced to approximately one step (1 cm-1). 50
The reductions were first attempted with an absolute flux distri bution of Vega calculated with ATLAS and the model atmosphere parameters of Schild, Peterson, and Oke (1971). The spectrum was computed in non-
LTE and included hydrogen line absorption due to Brackett a and
Humphreys 14. The calculated fluxes were placed on a magnitude scale per unit wavelength and convolved with the scanner bandpass to obtain the standard values needed to initiate the reductions, which proceeded as described in Section 2.4 until a total of 11 stars had been tied together as standards.
The reduction made in this manner proved to be unsatisfactory for two reasons. First, since a Lyr is rather faint at 4 pm, all of the measured scans for this star are noisy. Since the point-to-point scatter in the a Lyr data was not sufficiently averaged out in the small number of available observations, it reappeared in the calculated trans formation coefficients and hence also in all of the reduced spectra.
Second, the calculated transformation coefficients, besides being noisy, showed a wavelength dependence that was not smooth. This effect could not be attributed to telluric absorption, which is relatively small in this region (Chapter III), and it was concluded that the calculated energy distribution for Vega was in error in the regions of the hydrogen lines. An alternative approach was therefore tried.
Instead of basing the reductions on a model energy distribution for a Lyr, it was decided to try assuming blackbody energy distributions for the three warmest standard stars: 3 Gem (KO Illb), a Ari (K2 Illab), and a Boo (K2 IIIp). These stars are of relatively early type, bright 51
at 4 ym, and no spectral features were apparent in the raw data. In addition, synthetic spectrum calculations (Section 4.8) indicated that stars in this temperature range have featureless 4 ym spectra at the resolution of the scanner with neither SiO bands nor hydrogen lines present. The color temperatures measured on the Wing 8-color system were adopted. Since 4 ym is several times the wavelength of the energy maximum for these stars, the computed blackbody energy distribu tions are closely approximated by the Rayleigh-Jeans law (FAaX-1*) and are insensitive to the adopted temperature (see Table 9 below). To account for the relative brightness of g Gem, a Ari, and a Boo, a constant was added to the blackbody fluxes of each star in the standard deck to minimize the residuals in the scanner data. The remaining 8 standards were then added to the standard deck and the final values of the standard stars were determined by iteration. The final fluxes were scaled so that the magnitude of a Lyr at 3.996 ym was 0.0.
At this point the wavelength dependence of the transformation coefficients was again examined and was found not to show any features, although small (0.01 mag) point-to-point differences existed because of noise in the standard star data. To reduce further the scatter in the transformation data, a gaussian of full half-width equal to the bandpass o (88 A) was convolved with the measured transformation coefficients.
This procedure preserves the wavelength dependence of the transformation data and decreases the point-to-point scatter. Smoothed transformation coefficients were now used to iterate the standard deck. The final values for the program stars were computed from a self-consistent set 52
of standard star fluxes and smoothed transformation coefficients. The smoothed transformation coefficients obtained on 26 October 1977, 28
October 1977, and 8 November 1977 have been plotted against wavelength in Figure 5.
Section 4.4 - Derived Quantities
Blackbody color temperatures were computed from a two point index formed from the 8-color magnitude at 1(104) and the observed scanner flux at 3.996 pm. The latter wavelength lies shortward of the onset of SiO absorption and represents a point of minimum absorption in most of the stars observed. Continuum magnitudes measured at 3.996 pm
(averaged over three scanner data points) will henceforth be referred to as L(400). The measured fluxes were transformed to an absolute flux scale by the procedure described in Section 2.4. The I (104)-L(400) color index has been used in the study of infrared colors (Chapter V ) .
Equivalent widths of both the (2,0) and the (3,1) 2 8 Si160 bands have been derived. Since in many stars SiO absorption occurs at all program wavelengths longward of the (2 ,0 ) bandhead, the continuum level had to be extrapolated into the SiO-depressed region. The continuum level was calculated using L(400) as a continuum point and the 1(104)-
L(400) color index to derive a blackbody slope. For temperatures warmer than about 3000 K the computed continuum slope is insensitive to errors in the blackbody temperature caused by reddening and photometric errors since the blackbody distributions in this region can be closely approxi mated by the Rayleigh-Jeans law (FAaA-t+). MAG 26 OCT 77
8 NOV 77 1 ' BANDPASS
a i 8______» » " « » 4.00 4.05
Figure 5. Smoothed 4 )Jm transformation coefficients obtained on 28 October 77 (upper), 26 Oct 77 (middle), and 8 Nov 77 (lower) plotted vs. wavelength in air microns. The transformation coefficients are expressed on a magnitude scale per unit wavelength and have been smoothed using a Gaussian of half-width equal to the resolution (88 54
The (2,0) band equivalent width was computed by integrating between the wavelengths 3.998 and 4.016 ym, and the (3,1) band equivalent width was integrated from 4.038 to 4.058 ym. The total Si0 equivalent width
W(SiO) was defined to be the sum of these two terms.
Errors in the measured values of W(SiO) can result from errors in the photometry at the wavelengths used to integrate the SiO equivalent width and from errors in the computed continuum level in this region.
Since the blackbody slope has been shown to be insensitive to the adopted temperature for most stars, the largest source of error in determining the continuum is likely to be the L(400) magnitude. This wavelength has been assumed to be free of stellar blanketing in all stars. Both the synthetic spectrum calculations (Section 4.8) and the
Kitt Peak FTS spectra of a Ori, a Tau, and y Gem indicate this assump tion is likely to be valid for most stars, but it is possible that water vapor or some unknown absorber could depress both the L(400) continuum point and the SiO region in the cooler stars or in peculiar objects.
Since H 2O absorption occurs in the 2 ym region in giants later than M 6
(Baldwin, Frogel, and Persson, 1973), the measured values of W(SiO) are likely to be less accurate because of water vapor absorption in the coolest stars. The uncertainty in the measured value of W(SiO) resulting from errors in the photometry can be estimated from the repeated measurement of the standard stars. The probable error of 0 W(SiO) for the standards is 5 A.
The 4 ym scanner results for all stars except Mira variables are summarized in Table 7. The first three columns are self-explanatory. 55
Table 7
4 ym Scanner Results for Non-Miras
# of W(SiO) HD HR=BS Name Spectrum Scans L(400) A Notes
3627 165 6 And K3 III 1 0.340 2.8
3712 168 a Cas K0- Illb 1 -0.342 -2.4 1
6860 337 6 And M0 Ilia 1 -2.033 21.4
11961 M5 III 1 -0.090 46.7
12929 617 a Ari K2 Illab 4 -0.748 1.1
14270 AD Per M2.5 lab 1 1,744 25.6 2
14469 SU Per M3-M4 lab 1 1.103 30.6 2
14488 RS Per M4.5 lab 1 1.270 36.1 2
14528 S Per M4e la 1 0 .010 -28.5 2,3
17506 834 r) Per K3 Ib-IIa 1 0.003 3.8
18191 867 45 RZ Ari M 6- III: 1 -1.270 40.0
18884 911 a Get Ml.5 III 6 -1.884 25.1
19058 921 p Per M4 Ilb-IIIa 1 -2.213 31.1
20797 1009 M0 II 1 0.246 23.0
22649 1105 S5,3 1 -0.071 28.3
23475 1155 M2+ Ilab 1 -0.867 32.4
25025 1231 y Eri M0 III 1 -1.125 18.1
29139 1457 a Tau K5 III 10 -3.020 15.4
30959 1556 o 1 Ori M3S 1 -0.787 29.7
32068 1612 C Aur K5 II + B 1 -0.039 13.9
GP Ori SC 1 1.699 4.7 1 56
Table 7
4 ym Scanner Results for Non-Miras
(cont'd.)
# of W(SiO) HD HR=BS Name Spectrum Scans L(400) A Notes
32736 1648 W Ori C5 s 3 1 -0.927 7.9
36389 1845 119 CE Tau M2 Iab-Ib 1 -1.228 39.3
38944 2011 O Aur M1+ Ilia 1 0.440 21.8
39801 2061 a Ori M1-M2 Ia-Ib 3 -4.519 40.9
40239 2091 7T Aur M3 II 1 -1.069 40.4
42475 2190 TV Gem M0-M1 lab 1 0.569 43.9
42543 2197 6 BU Gem Ml-M2 Ia-Iab 1 0.482 18.4
42995 2216 n Gem M3 III 1 -1.654 31.6
44478 2286 y Gem M3 Illab 1 -2.106 29.5
44537 2289 ip1 Aur K5-M0 Iab-Ib 1 0.250 24.7
55383 2717 51 BQ Gem M4 III 1 -0.383 30.9
58061 VY CMa M 1 -3.172 -24.0
60414 2902 KQ Pup M2 Iabep+B 1 -0.254 37.0
60522 2905 U Gem M0 III 1 0.103 19.4
62509 2990 6 Gem K0 Illb 4 -1.174 -1.0
69267 3249 e Cnc K4 III 1 -0.005 10.9
76827 3576 p UMa M3 Illb 1 0.217 28.2
78712 3639 RS Cnc M 6e(S) 1 -1.979 32.4
80493 3705 a Lyn K7 Illab 2 -0.772 14.4
85503 3905 y Leo K1.5 III CN1 1 1.146 0.5 57
Table 7
4 pm Scanner Results for Non-Miras
(cont'd.)
# of W(SiO) HD HR=BS Name Spectrum Scans L (400) A Notes
86663 3950 TT Leo M2- Illab 1 0.345 — 5
108849 BK Vir M7- 1 -1.298 36.2 1
124897 5340 a Boo K2 IIIp 5 -3.128 1.7
126327 RX Boo M 8 2 -2.407 39.6
146051 6056 6 Oph M0.5 III 2 -1.423 19.7
156014 6406 a 1 Her M5 Ib-II 2 -3.785 40.8
161096 6603 3 Oph K2 III 1 0.126 2.5
165674 VX Sgr M4e 1 -1.842 —
172167 7001 a Lyr A0 V 4 -0.003 15.5 4
172380 7009 XY Lyr M4-5 II 1 -0.505 49.0
175865 7157 R Lyr M5 III 1 -2.319 39,1
186791 7525 Y Aql K3 II 3 -0.759 8.5
RW Cyg M3-M4 Ia-Iab 2 -0.021 20.2
196610 7886 EU Del M 6 III 1 -1.393 38.3
NML Cyg M 6 1 -1.874 —
206778 8308 £ Peg K2 lb 2 -0.977 6.8
206936 8316 y Cep M2 la 4 -2.292 18.9
207076 M7 III: 1 -1.907 27.3
208816 8383 W Cep M2 Ia-Iab 1 -0.435 22.4
212466 RW Cep K 0 O-Ia 2 1.461 -18.3 3 58
Table 7
A ym Scanner Results for Non-Miras
# of W(SiO) HD HR=BS Name Spectrum Scans L(AOO) A Notes
21A665 8621 MA III 1 -0.A12 3A.6
2169A6 8726 MO- lb 1 0.58A 29.1
217A76 8752 GO la 1 1.126 0.6
217906 8775 3 Peg M2.5 II-III 3 -2.AA3 22.6
221615 89A0 71 Peg M5- Ilia 1 -0.A27 36.6
Remarks concerning stars in Table
1 One or more noisy scans. Somewhat higher errors.
2 Supergiants near the h and x P®r cluster.
3 SiO in emission?
A Calculated value of W(SiO) not meaningful.
5 Missing data. W(SiO) could not be calculated. 59
Column 4 is the published spectral type repeated from Table 2 in Chapter
II. Column 5 lists the number of scans made for each star. Column 6 contains the mean observed scanner flux at 3.996 pm, on a magnitude scale normalized so that L(400) for a Lyr is 0.0. Column 7 is the average observed SiO equivalent width in Angstroms. The last column refers to notes which immediately follow the table.
Section 4.5 - Normal Oxygen-rich Giants and Supergiants
In Figure 6 the total SiO equivalent width W(SiO) is plotted against MK spectral type for non-variable and low-amplitude variable stars of spectral types K and M. In the figure luminosity class II and
III stars are shown as solid circles while supergiants of types lab and lb are indicated with open circles. Solid triangles refer to stars classified as la or Ia-Iab on the MI system. MS stars are indicated with a plus (+) symbol.
The relationship between band equivalent width and spectral type is well-defined for the giant and bright giant stars. The SiO bands have no appreciable strength in the early K stars, but they are defi nitely present at K5 and continue to increase in strength until at least M 6 . The data can be satisfactorily represented by a straight line, which has been drawn in Figure 6 . The two MS stars observed, o 1
Ori (M3S) and RS Cnc (M6 S), have SiO strengths similar to normal giants of the same spectral type. The mean giant and bright giant relation is listed in Table 8 . The effective temperature scale of Ridgway, et al. 6 0
50
TV Gam
40 119 To
30
W(SiO) O A AD Per
20 ERROR
KO Kl K2 K3 K4 K5 MO Ml M2 M3 M4 M5 MS M7 MS MK SPECTRAL TYPE
Figure 6. The relation between W(SiO) and MK spectral type for K and M stars. The peculiar supergiants discussed in Section 4.6 and Mira variables have been excluded. Luminosity class II and III stars are shown as solid circles, lab and lb super giants as open circles, la and Ia-Iab supergiants as solid triangles, and MS stars as plus (+) symbols. The solid line represents the mean relation for luminosity class II and III stars. 61
Table 8
Mean Relation between W(SiO) and MK Spectral Type for Giant and Bright Giant Stars
T W(SiO) MK eff Spectral Type (Ridgway et al. 1980)
K1 4610 0.0
K2 4450 2.0
K3 4270 6.0
K4 4095 10.0
K5 3980 14.0
MO 3895 18.0
Ml 3810 22.0
M2 3730 26.0
M3 3640 30.0
M4 3560 34.0
M5 3420 38.0
M6 3250 42.0 62
(1980) has been included in the table to facilitate comparisons with the model atmosphere calculations in Section 4.8.
A sequence of scans of K and M giant stars is shown in Figures 7 and 8 . Each scan begins and ends at slightly different wavelength points because of shifts in the wavelength scale caused by flexure
(see Section 4.3). The nearly straight line with each spectrum is a blackbody curve of the temperature corresponding to the I(104)-L(400) color and passes through the mean of the data points used to define
L(400).
In B Gem the SiO bands are not apparent and the spectrum is fea tureless apart from noise. In K5 to M2 stars the SiO bands are defi nitely present but sufficiently weak that the spectrum nearly reaches the continuum just shortward of the (3,1) head. In cooler stars the bands are much stronger, and all points longward of the (2 ,0 ) head are significantly depressed. It is noteworthy that in all spectra the (3,1) band is stronger than the (2,0) band. This is in agreement with synthetic spectra calculations (Section 4.8).
For the supergiants it is evident from Figure 6 there is a consid erable dispersion in SiO strength at a given MK type. Three of the supergiants identified in Figure 6 , 119 Tau (M2 Iab-Ib), a. Ori (M1-M2
Ia-Ib), and TV Gem (M0-M1 lab), have SiO equivalent widths nearly twice as large as giant stars of the same spectral type. Other M supergiants, notably RS Per (M4.5 lab), AD Per (M2.5 lab), and SU Per (M3-M4 lab) of the h and x Per cluster, have SiO strengths similar to those on the mean giant relation. The four la or Ia-Iab M supergiants observed, RW Cyg 63
. BETA GEM KO II IB >1860
1(400) ALPHA TAU 1-1 K5 III 3618
. DELTA OPH
MAG
12,0)
(3, n
.00 U.05 WAVELENGTH (MICRONS) Figure 7. Scanner observations of 6 Gem (KO Illb), a Tau (K5 III), and 6 Oph (MO.5 III). The MIC spectral types are indicated below the star name as well as the calculated I(104)-L(400) blackbody temperature. The wavelength scale is in air microns. The fImres are on a magnitude scale per unit wavelength subject to arbitrary normalization. The band pass (88 $)„ an 0.1 mag flux intervals and the positions of the OH and SiO features are shown below. Each scan is accompanied by a blackbody curve (solid line) of the temperature corresponding to the I(104)-L(400) color. The wavelength interval used to define the L(400) magnitude and to set the height of the blackbody curve is shown above the spectrum of a Tau. . BETfi PEG
3262
XT LYR
3052
. RX BOO
2563
■V' HRG
BRNDPfiSS
a U U *814.05 • U J WAVELENGTH (MICRONS)
Figure 8. Scanner observations of B Peg (M2.5 II-III), XY Lyr (M4-M5 II), and RX Boo (M8:e) are shovm in the same format as Fig. 7. 65
(M3-M4 Ia-Iab), p Cep (M2 la), W Cep (M2 Ia-Iab), and BU Gem (M1-M2
Ia-Iab), ail lie below the mean giant relation. Supergiant scans of three stars are shown in Figure 9.
Although a large variation of SiO strength among supergiants of the same spectral type is indicated by the data, further measurements are needed to show whether this is caused by actual differences in SiO abundance. Observations over a wider wavelength region or at higher resolution are needed to check that the region just shortward of the
(2,0) SiO bandhead is a clean continuum region in all stars, as has been assumed. Also, higher resolution is required to detect any emis sion which could fill in and weaken the bands as observed with the scanner.
Section 4.6 - Observations of Peculiar Supergiants
Five unusual, high-luminosity supergiant stars were observed on the program. Scans of these objects are reproduced in Figures 10 and 11.
S Per. The 8-color spectral type at the time of observation was
M4.7 la. The 4 pm data show a sharp rise in flux at the location of both the (2,0) and the (3,1) bandheads suggesting that SiO was in emis sion at the time of observation. Emission in the 4 pm SiO bands has been previously suspected by Merrill and Stein (1976), whose low-reso- lution measurements at two epochs indicated a sharp upturn in the spectrum at the location of the SiO features. Vibrationally excited
SiO has been detected in emission in the microwave region (Kaifu, Buhl, and Snyder, 1975). . VV CEP M2 IA-IAB 3091 ECL. BIN.
119 CE TAU M2 IAB-IB 3066
. RW CYG H3-U IA-IAB 2386
HAG
U.00 H.05 WAVELENGTH (MICRONS)
Figure 9. Scanner observations of the M supergiants W Cep (M2 Iaep), 119 CE Tau (M2 Iab-Ib), and RW Cyg (M3-M4 Ia-Iab) are shown on the Bame format as Fig. 7. . S PER M4E m 2238 H £ CHI
RW CEP KO O-IR 3396
MRC
(2,0)
BflNDPfiSS
U.00 WRVELENGTH (MICRONS)
Figure 10, Scanner observations of the very luminous supergiants S Per (M4e la) and RW Cep (KO O-Ia) are shown on the same format as Fig. 7. SiO emission may be present at the location of the SiO (2,0) and (3,1) bands. NML CYG
. VY CHfl H I 1422
HRG
12,0) BfiNDPfiSS
WAVELENGTH (MICRONS)
Figure 11. Scanner observations of the peculiar M supergiants NML Cyg, VX Sgr, and VY CMa are shown on the same format as Fig. 7. 69
RW Cep. The scanner data for this object also suggest that the
(2,0) and (3,1) bands of SiO are in emission. This result, if confirmed, is surprising in view of the early spectral type of this object (KO O-Ia).
However, it is possible that the emission at the SiO bands is only a feature of the most luminous supergiants since both S Per and RW Cep are generally considered to be among the most luminous late-type stars known. RW Cep is also well known for its abundance of emission lines in the violet (Merrill and Wilson, 1956) and a large infrared excess (cf.
Dyck, et al., 1971). Again, observations at higher resolution are needed to confirm the presence of SiO emission. Also because the spectrum is peculiar and quite noisy, it is difficult to know that the wavelength scale is correct.
VX Sgr, NML Cyg, and VY CMa. All three of these objects are well known for their excess radiation at infrared wavelengths (cf. Merrill,
1977). The present observations at 4 pm indicate that the region scanned is void of stellar absorption or emission features. The peculiar energy distributions observed strongly suggest that any photospheric SiO pre sent in these stars has been veiled by circumstellar emission. This result is consistent with the 4-5 pm observations of Geballe, Wollman, and Rank (1973) and Wollman, et al. (1973). No photospheric 4.7 pm CO lines were observed in VY CMa or NML Cyg despite strong first-overtone
CO absorption at 2.3 pm. The SiO (3,1) band was absent in their data for both VX Sgr and NML Cyg.
Although the blackbody continua shown in Figures 7 to 9 all have about the same slope, this is no longer true at very cool temperatures. 70
Table 9 shows how the blackbody magnitude changes over the width of the
interval scanned as a function of temperature. Only at temperatures below 1500 K does the slope show noticeable change. For temperatures
less than 720 K the blackbody flux increases to longer wavelengths.
Blackbody fits to the general trend in the spectra yield temperatures of about 700 K, 1500 K, and 400 K for VX Sgr, NML Cyg, and VY CMa, respectively. In Figure 11 the blackbody curve indicated for VY CMa is based on the simultaneous measurement of 1(104). The difference between the observed and calculated 4 ym slopes clearly documents the presence of excess emission in this spectral region. No blackbody curves accompany NML Cyg and VX Sgr because simultaneous 8-color photometry could not be obtained.
The nature of the infrared flux distributions of highly luminous cool supergiants has been the subject of much controversy. Humphreys
(1974) and Gilman (1974) have suggested that the reported weakening of absorption lines in the near infrared is due to chromospheric H- bound- free emission, and that optically thin free-free radiation is responsi ble for the observed 3-8 ym energy distributions. The results of
Fawley (1977) indicate that the identification of H- bound-free emission is erroneous and that the excess can be explained by overcorrections for interstellar extinction and TiO opacity. Fawley identifies the 3-8 ym excess with thermal reradiation from the circumstellar shell. The presence of thermal circumstellar water vapor emission (Tsuji, 1978b) may further complicate the interpretation of the observed energy distri butions in the 5-8 ym region. Table 9
Blackbody Indices in the 4 ym Region
THETA TEMPERATURE (K) L(400)-L(406)
0.1 50400 0.065
1.0 5040 0.061
1.5 3360 0.056
2.0 2520 0.052
2.5 2016 0.047
3.0 1680 0.043
5.0 1008 0.022
10.0 504 -0.035
20.0 252 -0.151
L(400) = magnitude per unit wavelength at 3.9965 ym
L(406) = magnitude per unit wavelength at 4.0576 ym 72
Section 4.7 - Observations of Mira Variables
The observations of Hinkle, et al. (1976) indicate that the SiO
strengths in Mira variables vary markedly with phase. Unfortunately,
it was not possible to follow individual stars with phase on this pro
gram. A total of ten Miras were observed, two of which were scanned a
second time after an interval of about one year. Although the behavior with respect to phase is not well defined in these data, the strongest
SiO absorption was observed in stars near minimum light while SiO was weak or absent in stars near maximum light. Both of these results
are consistent with Hinkle, et al. (1976).
Scans of the Mira variables are presented in Figures 12 to 15. The
I(104)-L(400) blackbody temperatures and calculated 4 ym continuum levels
have been indicated. In four cases the simultaneous 1(104) magnitudes
observed with the 0.4-m Kitt Peak telescope have been used in this cal
culation. For the other SiO observations, 1(104) magnitudes have been
estimated from published infrared light curves (Lockwood and Wing, 1971;
Wing and Lockwood, 1973) and unpublished scanner (Wing, 1967a) and 8-
color data.
Most of the Miras observed show rather strong absorption at the
positions of both SiO bands so that there can be little doubt regarding
the identification of this molecule. However, some of the spectra have
a different character. In some cases the SiO bands seem to be absent,
and it's not clear at this resolution if this is due to a deficiency
of SiO molecules in the atmosphere, or the presence of additional absorp
tion in this region (e.g.^O), or to a filling in of the SiO features . R RND S6.6E 2128 3444.852
0M1CRON CET GM6E 2802 ’ 3142.959
IK TRU H8.0-H10 1364 ’ 3417.911
13,1)
4.05 WAVELENGTH (MICRONS)
Figure 12. Scanner observations of the Mira variables R And (phase 0.77), o Cet (phase 0.88), and IK Tau (phase 0,5) are shown on the same format as Fig. 7. U OR I M8E 2572 3455.942
R CNC M6E-H8E 2811 3456.060
. U ROL S3.9E
3036.727
OH
BflNDPRSS
U.00■ UU 4.05 1•UJ WAVELENGTH (MICRONS)
Figure 13. Scanner observations of the Mira variables R Aur (phase 0.79), U Ori (phase 0.16), R Cnc (phase 0.87), and W Aql phase 0.20) are shown on the same format as Fig. 7. R C T G S3.9E-S6.8E 2070 - 3447.675 '1
. CHI CTG S7,1E-S10 2325 3036.745
DH
WAVELENGTH (MICRONS)
Figure 14. Scanner observations of the Mira variables R Cyg and x Cyg are shown on the same format as Fig. 7. Upper observation was obtained at phase 0.34, middle at 0.31, and lower at 0.17. . CHI CTG S7.1E-S10.1E 2268 34414.699
R CflS M6E-H8E 1899 3036.990
. R CHS M6E-H8E 1971 3 W 2 . 9 3 1
HRG
(2.0)
U.00U.00 11.05 WAVELENGTH (MICRONS)
Figure 15. Scanner observations of the Mira variables x Cyg (phase 0.17) and R Cas (phase 0.68 and 0.63) are shown on the same format as Fig. 7. 77 by emission by the molecule itself. Furthermore, the hydrogen line Bae which is nearly coincident with the location of the (3,1) head, may appear in emission at phases near maximum light. At high resolution
SiO emission has been noted in Miras (Ridgway, Hall, and Carbon, 1977), but it is not known if this is a common occurrence. Line doubling has been noted for CO (Hinkle, 1978) and other features and may also occur for SiO at some phases, further complicating the interpretation of the
SiO band strengths.
The observations are summarized in Table 10. The first three columns contain the HD number, the HR=BS number, and the variable star name. Column 4 is the spectral type determined by simultaneous 8-color photometry. Column 5 is the Julian date (+2440000) of observation.
Column 6 contains the phase counted from visual maximum. Phases reported here are based on observations made by the American Association of Vari able Star Observers (Mattei, 1978) except for IK Tau for which the elements of Wing and Lockwood (1973) were used. Columns 7 and 8 are the 1(104) and L(400) magnitudes derived from 8-color and scanner data, respectively. The calculated 8-color and I(104)-L(400) color tempera tures (discussed in Section 5.3) are in columns 9 and 10. For stars which do not have simultaneous 8-color data, the magnitudes and color temperatures in columns 7 and 10 are enclosed in parentheses. The final column contains the calculated SiO equivalent width in Angstroms.
These values have been calculated in the same manner as for the non-Mira stars. Because of the peculiarities of the 4 ym spectra sometimes observed and because of the problems of interpretation rioted earlier, Table 10
4 vim Scanner Results for Miras
8-color W(SiO) HD HR=BS Name Spectrum JD Phase 1(104) L(400) T(8-c) T(l-4ym) A
1967 90 R And 3444.85 0.77 (3.2) -0.279 (2128) 19.3
14386 681 o Cet M 6.8 3442.96 0.88 -1.087 -3.434 2659 2802 (-11.5)
IK Tau (M10) 3447.91 0. 5±0.1 (4.1) -2.050 (1384) 59.5
34019 1707 R Aur 3455.99 0.79 (1.52) -1.313 (2462) 15.0
39816 2063 U Ori 3455.94 0.16 (0.95) -1.711 (2572) 17.6
69243 3248 R Cnc 3456.06 0.87 (1 .2 ) -1.136 (2811) -2.5
W Aql 3036.73 0.20 -1.705
185456 R Cyg 3034.72 0.34 (3.8) 0.283 (2 1 2 0 ) 74.7 185456 R Cyg 3447.68 0.31 3.780 0.165 1562 2070 76.3
187796 7564 X Cyg 3036.75 0.17 (0 .0 ) -3.112 (2325) 58.5 187796 7564 X Cyg M7.7 3444.70 0.17 -0.038 -3.221 1946 2268 34.7
224490 9066 R Cas (M10) 3036.99 0.68 (1.4) -2.615 (1899) 38.2 224490 9066 R Cas M9.5 3442.93 0.63 1.452 -2.414 1384 1971 42.1 79
the calculated values of W(SiO) may not be meaningful. In the case of o Cet the spectrum is so obviously peculiar that the value of W(SiO) has been enclosed in parentheses.
Individual objects are discussed in the following paragraphs.
R And. A single 4 ym scan of this strong S star was obtained during rising light. Both the (2,0) and (3,1) bands are clearly visible but only of a strength comparable to a K5 giant star.
o Cet. A single observation was made shortly before maximum light, when the 8 -color spectral type was M 6 .8 . The spectrum is peculiar with no identifiable features. The excess above the continuum level in the longward half of the spectral region may be due in part to Ba emission.
IK Tau (NML Tau) . Strong SiO absorption from both the (2,0) and
(3,1) bands is evident in a scan obtained at phase 0.5+ 0.1. IK Tau was invisible at the 1.3-m, and unfortunately it was impossible to obtain a simultaneous classification from 8 -color photometry with the
0.4-m telescope. However, since IK Tau is the only star known to reach spectral type M10 consistently at every minimum, it may be assumed that the type was at or near M10 when this SiO scan was made.
An 1(104) magnitude of 4.1 has been estimated from the light curves of Wing and Lockwood (1973).
R Aur. A scanner observation obtained in November 1977 on the rising branch of the light curve shows a decrease in flux at the posi tion of the (2,0) bandhead. However, the (3,1) head is not visible in the scan. Perhaps the (3,1) band has been filled in by Ba emission. 80
Clearly, higher resolution measurements are required to interpret the spectrum properly.
U Ori. A scan obtained shortly after maximum light is very similar to the R Aur observation. The (2,0) bandhead is present, but the (3,1) bandhead is not.
R Cnc. A single observation of this star was obtained approaching maximum light. The (2,0) bandhead is absent and an emission feature, probably Ba, is observable near the location of the (3,1) bandhead.
W Aql. This nearly pure S-type Mira was observed shortly after maximum light. No features are visible in the spectrum.
R Cyg. This star, another nearly pure S-type Mira, exhibited the
strongest SiO bands observed. Observations were obtained in September
1976 at phase 0.34 and in October 1977 at phase 0.31. Although the
1976 scan is noisier, both observations show similar SiO strengths.
R Cyg was more than 0.1 mag fainter at 4 pm during the latter observa
tion .
X Cyg. Observations of this mild S star in September 1976 and
October 1977 were both made at phase 0.17. The (2,0) and (3,1) bands
are visible in both spectra but are weaker in the 1977 data. As in
the case of TiO absorption in Miras (Wing, 1967a), the strength of the
SiO features at a given phase may vary from cycle to cycle. At L(400),
X Cyg was more than 0.1 mag brighter at the latter epoch.
R Cas. Observations near minimum light (phase 0.68 and 0.63,
respectively) were made in September 1976 and October 1977. The
spectral type of this star at normal minimum has been adopted as the 81
definition of type M10 (Wing and Lockwood, 1973), and the type at the
time of the 1977 observation was M9.5. The (2,0) and (3,1) bands are
strong at both epochs. In addition, an absorption feature is visible
in both spectra near the location of the 2^SiO bandhead. High resolu
tion measurements are needed to confirm this identification and to
derive silicon isotope ratios. L(A00) differed by 0.2 mag at the two
epochs.
Although only a relatively small number of observations were
obtained some trends are evident. Among the M-type Miras, the strongest
SiO bands were observed at the latest spectral types, which occur near minimum light. Near maximum light the bands are considerably weaker,
and in fact are much weaker than normal M giant stars of the same TiO
types (about M6 ) . A similar effect is observed in S-type stars whose
SiO strength varied from vanishingly small to among the strongest ob
served. Any systematic differences between the M and S stars are
masked by the large changes with phase displayed by both.
Since the two stars of latest type, R Cas and IK Tau, show strong
SiO bands, it is evident that SiO absorption dominates over anything
else in this region at the lowest temperature. This result would
appear to argue against the idea that the SiO features in warmer Miras
are obliterated by absorption, by water vapor, or some other temperature
sensitive molecule. Likewise, since the nearly pure S star R Cyg,
whose spectrum is full of unidentified bands in other regions, shows
strong SiO and no evidence for other absorbers, it seems unlikely
that the weakness of the SiO bands in other S stars (e.g.W Aql) can 82
be explained by absorption by other molecules. Thus, this indicates that there are no strong absorbers in this spectral region other than
SiO in stars of any type. It seems likely, therefore, that the peculiar shapes of the spectra of some Miras are more the result of emission by
SiO and/or Ba, rather than absorption by other molecules.
Section 4.8 - Computation of Synthetic 4 pm Spectra
The synthetic spectrum program MOOG was used to generate theoreti cal spectra to compare to the observed 4 pm scans. The line list in cluded the first-overtone SiO bands, the bands of the Av=-3 sequence of the CN red (A2 tt-X2 Z) system, the fundamental vibration-rotation bands of OH, and atomic lines. The computation of the synthetic spectra is discussed in this section.
The vibrational and rotational constants used for the calculation of the SiO line positions were taken from Table 1 of Beer, Lambert, and
Sneden (1974). These values were obtained from microwave data and the observed first-overtone line positions in a Ori. The standard isotopic relations (Herzberg, 1950) were used in the computation of the line positions for 29SiO and 3 0 SiO.
The transition probabilities for SiO adopted here are those calcu lated by Hedelund and Lambert (1972). These values are based on a fit to the electronic dipole moment measurements of Raymonda, Muenter, and
Klemperer (1970). A Morse potential function was assumed. The rota tional line oscillator strengths have been computed on the assumption of negligible vibration-rotation interaction. Beer, Lambert, and 83
Sneden (1974) have shown that the effect of vibration-rotation inter action is small for the first-overtone lines. Isotopic lines were
assumed to have the same gf values as the corresponding lines of 2 8 SiO.
The OH fundamental-band line positions were calculated from the
constants derived by Maillard, Chauville, and Mantz (1976) with a computer program furnished by D. L. Albritten (see Albritten, et_ al.,
1973, for a description of the program). A set1 of isotopic relations, also furnished by Albritten (1978) , was used to compute line positions
for 170H and 180H. The accuracy of these positions is unknown since
the isotopic lines have not been observed in the laboratory.
The experimental OH oscillator strengths of Roux, d'Incan, and
Cerny (1973) were used for the (1,0) and (2,1) bands. These values were determined from spectra of an oxyacetylene flame in thermal equili brium. In the calculations the rotational line strength formulae of
Kovacs (1969) were used and the experimental polynomial fits for the
Herman-Wallis factor were adopted. Since absolute strength measure ments have not been reported for the higher vibrational levels of this
sequence, the relative transition probabilities of Murphy (1971) were
scaled to the absolute values of Roux, d ’Incan, and Cerny (1973) and
the Herman-Wallis factors were taken to be unity for these bands.
The OH oscillator strengths must be regarded as quite uncertain.
The adopted values differ considerably from the theoretical results
of Mies (1974). The observed transition probabilities for the (1,0)
atd (2 ,1 ) bands are approximately a factor of two smaller than the 84
theoretical values. Also, the theoretical values predict strong
vibration-rotation interaction for the fundamental bands.
The CN line positions have been calculated from the constants of
Fay, Marenin, and van Citters (1971). These values are based on the measurements of Davis and Phillips (1963) in the visible and near in
frared and cannot accurately reproduce the observed infrared line posi
tions of Cerny, et a l . (1978). However, the calculated positions are
accurate enough for the bandpass synthesis used in this study. Posi
tions for both 32CN and ^3CN lines were calculated.
For the oscillator strengths and dissociation energy of CN, I
have adopted f = 1.0 x 10-3 and D = 7.89 eV (Carbon, 1973). The 00 0 Franck-Condon factors of Spindler (1965) based on RKR potentials have
been used. The Hc5nl-London factors of Earls (1935) were normalized
via the rules of Tatum (1967).
The atomic line list and semi-empirical gf values of Kurucz and
Peytremann (1975) were used in the computations. The line positions
should be sufficiently accurate for bandpass synthesis.
The solar composition atmospheres of Johnson, Bernat, and Krupp
(1980) and Bell, et al. (1976a) were used to generate the synthetic
spectra. The abundances of the elements adopted were those of
Gustafsson, et al. (1975). Terrestrial values were used for the
isotope ratios of carbon, oxygen, and silicon. In the dissociation
equilibrium a total of 74 molecules were included. 85
A synthetic spectrum calculated with the Johnson, Bernat, and
Krupp (1980) solar-composition opacity-sampling model of effective temperature 3200 K and log g = 1.0 is shown in Figures 16 to 20. A microturbulent velocity of 2 km/s was adopted. Virtually all of the features calculated to be visible are due to either SiO or OH. The isotopic (2,0) bandheads of 29SiO and 30SiO can be weakly seen in the spectra at 2481.7 and 2467.2 cm-1, respectively. Both isotopic band heads have been seen in some late-type stars (Hinkle, et al., 1976;
Merrill and Ridgway, 1979).
The calculated spectra have been convolved with the spectrometer bandpass to produce synthetic 4 pm scans. A sequence of synthetic scans is shorn in Figure 21. It can be seen that the region is pre dicted to be featureless in warmer stars (T > 4500 K ) . This result e supports the use of smooth energy distributions for 6 Gem (KO Illb) , a Ari (K2 Illab), and a Boo (K2 IIIp) in the reduction procedure. Both the OH and SiO depressions are weakly present in the 3800 K synthetic scan, and the strengths of both molecules are predicted to increase towards lower effective temperatures. The depression near 3.985 pm is caused by about a dozen strong lines of OH as can be seen from a comparison of Figures 3 and 20. The OH depression occurs at the short- wavelength edge of the region scanned. The downturn in flux due to OH can be seen in some of the stellar data in this chapter and in Appendix
A.
To calculate SiO equivalent widths from the theoretical scans, a procedure similar to the method used to treat the observations was 8 6
WAVELENGTH IN ANGSTROMS <106.20 <10600 <10580 wdsbo <105.40 <10820 ATOM ATOM SIO SIO OH OH D.O
3 2 0 0 / 1 . 0 / S HAVENUHBEA CN CN
<10620 <10600 <10580 H0560 <10540 <10520 <10500 <10<160 <10<160 HPVELENDTH IN ANGSTROMS
Figure 16. A portion of a synthetic spectrum generated with MOOG. The model atmosphere used was an opacity-sampling log g = 1.0, Te = 3200 K model. Solar elemental abundances, terrestrial isotope ratios, and a microturbulent ve locity of 2 km/s were assumed in the calculations. The positions of atomic, SiO, OH, and CN lines are indicated above and below. Molecular lines of the terrestrially most abundant isotope are indicated with arrows while short lines mark features from less abundant isotopes. The ^ S i O bandhead at 2467.2 cm is weakly visible in the computed spectrum. 87
WAVELENGTH IN ANGSTROMS 40440 40420 40400 4O3B0 ----40360 -- ATOM ATOM
0H 0H 0.0 0.0
3300 / 1.0 /5 HAVENUHBER CN CN
40440 40420 40400 40380 40360 40340 40320 40300 402B0 WAVELENGTH IN ANGSTROMS
Figure 17. Continuation of the synthetic spectrum shown in the previous figure. The 28SiO (3,1) bandhead is at 2472.9 cm-l. The 28SiO bandhead at 2481.7 cm-^ is weakly visible in the computed spectrum. 8 8
WRVELENGTH IN RNGSTROMS 10260 102.10 10220 10200 MO 1.80 101,60 >101110 10120 RTOM IIII Ill 1 BI II II I II I II 1 I II I III I II II RTOM /N /Is /ts ''T' /tv/K/tv/^/tv/*v/f\/tv/ts./tv/N /^/ts/ts /N4v/K4v 'N't'* 'f'/tv / vK SIO SIO ' 1 N 7 \ 7 \y NKM7I 4*1' 44' OH OH ys , f w
3 2 0 0 / 1 . 0 / S sues1 ------M r HRVENUMBER A * s /txJ^m s /tXtv /tv /fv /TVKi^v /fv /fv /|v4»^\ /fvtxfvTtftv CN ^ ii inn CN VIW4I'' 4v 44* 1 1 1 1 I*/' 4444*1/
>10260 >102«10 10220 10200 101B0 10160 10110 10120 WRVELENGTH IN ANGSTROMS
Figure 18. Continuation of the synthetic spectrum shown in the previous figure. Most of the strong lines are due to the (2,0) band of 28SiO. 89
WRVELENGTH IN RNGSTROMS 40040 ------399,GO39980 RTOM 51 III RTOM
OH OH 0.0
3200/1.0/5 HRVENUMBER CN CN
40100 40080 40060 40040 40020 40000 399B0 39960 39940 WRVELENGTH IN RNGSTROMS
Figure 19. Continuation of the synthetic spectrum shown in the previous figure. The (2,0) 28SiO bandhead occurs at 2497.2 cm--*-. The region shortward of the (2,0) head is nearly free of line blanketing. 90
WAVELENGTH IN ANGSTROMS 39920 39900 39880 39860 398,40 398,20 396,00 39780 39760 RTOM RTOM SIO OH OH 0.0
3200/1.0/S 2SiBr"~ MAVENUMBER CN CN
39920 39900 39880 39860 39820 39800 39780 39760 HAVELENG N ANGSTROMS
Figure 20. Continuation of the synthetic spectrum shown in the previous figure. Most of the strong lines are due to OH. 91
4500/2.25 SOLAR 2 KM/SEC
3800/1. 0 SOLAR 2 KM/SEC
3600/1.0 SOLAR 2 KM/SEC
3400/1.0 SOLAR 2 KM/SEC
3200/1o 0 SOLAR 2 KM/SEC
3000/1.0 SOLAR 2 KM/SEC
OH
(2 . 0 )
BANDPASS
WAVELENGTH (MICRONS)
Figure 21. Synthetic k ym scanner spectra are shown on the same format as Fig. 7. All spectra have been computed with solar elemental abundances, terrestrial silicon, oxygen, and carbon isotope ratios, and a 2 lan/s micro- turbulent velocity. The effective temperatures and surface gravities of the models are indicated to the left of the spectra. The uppermost model is from the grid of Bell et al. (1976a). All other models are from Johnson, Bernat, and Krupp (1980). 92
adopted. For each model a synthetic color temperature was computed
from a blackbody fit to the continuum fluxes at 1(104) and L(400).
This color temperature was used then to calculate a blackbody continuum
in the SiO-depressed region. The SiO equivalent widths were computed by integrating the synthetic scans over the same wavelength intervals used for the observational data.
The calculated SiO equivalent widths are tabulated in Table 11.
All models assume solar elemental abundances and terrestrial isotope
ratios. For each model SiO equivalent widths have been computed for
a range of microturbulent velocities. The calculated values are com
pared to the observations in the next section.
Section 4.9 - Comparison between Observed and Synthetic Spectra
The observed line strengths are determined primarily by four
quantities: effective temperature, surface gravity, chemical composi
tion, and microturbulent velocity. It is impossible from low resolution
observations to determine all of these simultaneously. However, the
effective temperature can be determined independently from color measurements, and the models can be used to explore the effects of varying the other parameters. In this section solar-composition model atmospheres have been used to study the relation between W(SiO)
and effective temperature for various values of surface gravity and microturbulent velocity. The effect of varying the composition could not be studied since realistic M star models are only available for
solar composition (Johnson, Bernat, and Krupp, 1980). 93
Table 11
Synthetic SiO Equivalent Widths (£)
W(SdO) in £ E, = 2 km/s £ = 4 km/s E, = 7 km/s Teff (K) log g
4500 2.25 -0.4
4000 0.0 4.8 1.0 4.6 4.9 2.0 4.6
3800 0.0 11.9 1.0 12.2 14.0 2.0 10.6
3600 0.0 23.9 1.0 22.0 27.5 2.0 16.5
3400 0.0 34.0 44.0 1.0 28.6 37.2 2.0 20.4
3200 0.0 41.9 1.0 32.6 43.0 2.0 22.9
3000 0.0 45.5 62.6 1.0 34.6 45.8 2.0 24.7 94
In Figure 22 the effect of varying the surface gravity is shown for a constant value of the microturbulent velocity (2 km/s). Models with log g = 0.0 are denoted by "x", open circles represent models with log g = 1 .0 , and plus (+) symbols indicate models with log g = 2 .0 .
All models with the same value of log g are connected with dashed lines. The observed mean giant and bright giant relation is shown as a solid line. The temperature scale of Ridgway, et al. (1980) has been used to convert the abscissa of the observed relation from MK spectral type to effective temperature. The synthetic spectra predict a mono tonic increase in SiO strength to cooler temperature in accordance with the giant star data. For cooler stars the SiO bands are predicted to be stronger in stars of lower surface gravity.
The importance of the choice of the microturbulence velocity has been tested for models with a surface gravity of log g = 1.0. In
Figure 23 model results are shown for a microturbulent velocity of
2 km/s (open rectangles) and of 4 km/s (solid rectangles). At the warmer temperatures the synthetic SiO equivalent width is nearly inde pendent of microturbulent velocity, but it becomes increasingly sensi tive to microturbulence in the cooler models as the bands saturate.
The calculated SiO strengths are for most cases weaker than the observed values. It is not clear what is the cause of the discrepancy.
The possibilities include: (1) the Si and/or 0 abundances differ from the solar values, (2) the SiO oscillator strengths are in error,
(3) the models do not accurately represent the atmospheric structures, and (4) non-LTE effects are important. However, the general increase 95
g / u@e ®
v /
fU-JL-Jt—rt— {L_J— fl n n fl a I— 5 0 0 0
:CTIVE Ti
Figure 22. Plot of W(SiO) in Angstroms vs. effective temperature for solar-composition opacity—sampling models with different surface gravities and a fixed microturbulent velocity of 2 Ian/s. The positions of models with log g = 0.0, 1.0, and 2.0 are indicated with ' V , "o", and "+" symbols, respectively. Sets of models with the same surface gravities have been connected with dashed lines. The solid line is the mean giant and bright giant relation taken from Table 8. The effective temperature scale of Ridgway et_al. (1980) has been used to convert from spectral type to effective temperature. 96
rr=T-T-
\ f 2 (cm/s
ULU
’FECTBVE TEMPERATURE CK)
Figure 23. Plot of W(SiO) in Angstroms vs. effective temperature for solar-composition opacity-sampling models of log g = 1.0 showing the effect of microturbulence. Open rectangles indicate the results obtained with a microturbulent ve locity of 2 km/s while solid rectangles represent values computed xjith k lon/s. Sets of models with the same micro turbulent velocity have been connected with dashed lines. The solid line is the mean giant and bright giant relation taken from Figure 22. 97
in strength with decreasing temperature is at least qualitatively the same. The small scatter in the observed relation for normal giants
(as noted in Section 4.5) implies that for these stars there exists a relatively small range in the chemical and physical variables.
For the cooler stars an enhancement in SiO strength can be brought about by either lowering the surface gravity or increasing the microturbulence. Unfortunately, it is impossible to distinguish between these two effects on the basis of these measurements alone.
It will be important to study a sample of M stars at high resolution to determine the range of microturbulent velocity in the atmospheres.
At present, the interpretation of the results is hampered by a lack of knowledge of the atmospheric parameters of M stars. These uncertainties are largest in the coolest stars where the model atmos phere results are most sensitive to both log g and the microturbulent velocity. For this reason, additional scanner measurements of SiO strengths in the cooler stars would not be useful at the present time for deriving elemental abundances. Also, abundance differences among the cooler stars may be masked by intrinsic differences in these parameters. The observed differences in SiO strengths among the M supergiants could, in part, result from these effects. The SiO bands may prove to be very useful for abundance studies of warmer stars
(T > 3600 K) where the bands are weak and show little dependence on log g* CHAPTER V
INTERPRETATION OF INFRARED COLORS
Section 5.1 - Introduction
The problem of determining the temperatures of late-type stars is an important one and has been the subject of much recent work. The fundamental parameter of interest is the effective temperature, which can be obtained directly from the angular diameter and bolometric mag nitude of a star. The number of direct measurements of effective temperature is increasing rapidly, and it is important to be able to calibrate the measured effective temperatures in terms of one or more temperature-sensitive parameters which can be conveniently, accurately, and quickly obtained. A valid system of relative temperature indices can be used to interpolate between the effective temperature data, and the calibration can provide a means of determining the effective tempe’rature of other stars.
Color temperatures, obtained from a blackbody fit to the observed fluxes, expressed on an absolute scale, at any two wavelengths, are particularly useful for this purpose. However, color temperatures must be interpreted with care. Since the energy distribution of any star differs from a blackbody curve, different color temperatures will be obtained with different choices for the two wavelengths. Color
98 99 temperatures are affected by the line absorption and continuous opacity in both bandpasses. They are, of course, also affected by interstellar reddening.
Despite these problems, color temperatures have been frequently used to estimate the effective temperatures of stars. Ideally, any system established for this purpose should have the following charac teristics. First, the two wavelengths used to form the color index must be widely separated in wavelength so as to be sensitive to changes in the slope of the energy distribution with respect to temperature.
Second, since the spectra of most late-type stars are heavily blanketed in the visible region, bandpasses should be selected in the near infra red or infrared where continuum regions can be located. Narrow-band measurements are preferable, since even in the infrared, wide-band photometry cannot avoid molecular bands and atomic lines. The differ ence between the continuous opacities at the two wavelengths used to form the color index should be sufficiently small that the continuous fluxes will both originate from the same range of stellar layers. This will make the color temperatures insensitive to differences in the tem perature structure caused by composition variations among stars. Color temperatures obtained at wavelengths selected on the basis of these criteria will have values close to the effective temperature, and the remaining differences will be systematic and can be calibrated. The calibration of the color temperature in terms of the effective temper ature can be made either from direct measurements (angular diameters and bolometric magnitudes) or from calculations made with model atmos pheres. 100
A new effective temperature scale for luminosity class III K and
M stars has been determined from angular diameter measurements and infrared photometry (Ridgway, et al., 1980). The new scale assigns systematically higher temperatures to M giants than previous scales
(Johnson, 1966; Dyck, Lockwood, and Capps, 1974). The new calibration is given both in terms of the color temperature on ..the 8 -color system and also as a function of spectral type. According to the empirical results, the color temperatures measured on the 8 -color system are lower than the effective temperatures by 150 K at MO and by 650 K at M 6 .
The implication of this result is that the 0.7540 pm shortward continuum point is affected by blanketing which increases to later spectral types.
It is therefore of interest to test color temperatures that use differ ent continuum points to see if any of them are more direct indicators of the effective temperature.
Extensive grids of line-blanketed, flux-constant models (cf. Bell, et al., 1976a; Johnson, Bernat, and Krupp, 1980) are available and can be used to calculate synthetic energy distributions and colors for comparisons with observations (cf. Gustafsson and Bell, 1979). The synthetic colors can be compared with the measured values to test the accuracy of the model atmosphere predictions and to calibrate color indices in terms of fundamental parameters including the effective temperature. The model atmosphere results, therefore, provide an independent test of the effective temperature scale derived from angular diameter measurements. 101
Detailed comparisons between observations and model predictions have proved to be difficult. Wide-band colors are important but have certain disadvantages since they involve complex integrations of stellar and telluric features with the telescope-photometer-filter sensitivity function over large wavelength regions. Since numerous atomic and molecular lines invariably occur in any wide spectral region of a cool star, the interpretation of wide-band colors is complicated by uncertainties in the line data (missing lines, errors in the dissocia tion energies and gf values) and by the sensitivity of line blanketing to changes in temperature, surface gravity, chemical composition, and microturbulent velocity. In addition, systematic discrepancies may arise if LTE is assumed in computing the spectrum. Since all of these parameters affect wide-band colors, it is difficult to use them to determine accurate values of effective temperature, surface gravity, and chemical composition. Narrow-band colors, however, can be measured at points nearly free of blanketing so that the comparison between observed and calculated colors is not subject to the large uncertainties mentioned above.
In the present study narrow-band magnitudes were measured at four continuum points between 1.29 and 4.00 pm with the Kitt Peak grating spectrometer. For most of the stars observed, measurements are also available on the Wing 8 -color system so that color temperatures can be formed over an extended wavelength region. 102
Section 5.2 - Description of the Bandpasses
The purpose of the infrared colors program is to provide narrow band measurements of infrared fluxes in oxygen-rich giants and super giants at a series of widely spaced continuum points in the infrared.
To provide the widest possible wavelength coverage, data obtained with the Kitt Peak infrared scanner has been combined with measurements made with the Wing 8 -color system. The program wavelengths cover the interval from 0.7 to 4.0 pm.
The 8-color photometric system has been described in detail else where (Wing, 1971). The continuum points for this system are at 0.7540 pm, 0.7810 pm, and 1.0395 pm. The 0.7540 pm point is used as a con tinuum point for K4 to M 6 stars while the 0.7810 pm serves as a con tinuum point for G and K stars. In the 8 -color reduction procedure corrections are applied at 0.7540 pm for slight contamination by CN
(White and Wing, 1978). The 1.0395 pm bandpass is an excellent con tinuum point in a wide range of late-type stars, and absorption seldom exceeds 5% in any star (Wing, 1967b). The half power bandpasses for the 8 -color filters are 50 A (=0.005 pm), 40 R, and 50 R at 0.7540 pm,
0.7810 pm, and 1.0395 pm, respectively. For stars later than M 6 , V0 bands suddenly appear strongly at 0.7540 pm, contaminating the short- ward continuum point of the 8 -color system. The points measured with the Kitt Peak grating spectrometer (AUDREY) will now be discussed individually. 103
1.287 pm (7768 cm-1) . Magnitudes reported for this wavelength, referred to as J(129), represent an average of measurements made at three overlapping bandpasses centered at 1.286 pm, 1.287 pm, and 1.288 o pm, with an exit slot of 29 A (0.0029 pm) in third order. The effec- o tive bandpass of the resulting average is about 40 A. Inspection of high resolution spectra of a Ori, a Her, p Gem, and a Sco indicates this region is nearly free of both stellar and telluric features.
2.101 pm (4758 cm-1) . Magnitudes for this point, designated K(210), are the average value of measurements at five wavelengths between 2.098 pm (4766 cm-1) and 2.104 pm (4751 cm-1). Data were obtained in second o order (bandpass = 44 A ) , and the effective bandpass of the average is about 80 X. This region in e Tau is shown in Figure 24 and in a Ori in
Figure 25. For the a Ori data (solid line), a synthetic spectrum (plus symbols) computed with only CN lines is also shown. Line positions used are from Cerny, et al. (1978) and CN strengths were computed fol lowing Carbon (1973) . It can be seen that many of the weak lines ob served in this region are due to CN. In Table 12 the mean depression in magnitudes across the bandpass as determined from high-resolution spectra has been compared with the CN index of the 8 -color system.
The depressions are small but correlate well with the near-infrared
CN data. To correct the photometry for CN contamination to a first approximation, the following quantity has been added to the observed fluxes at this wavelength:
Absorption at 2.101 pm (mag) = 0.40 CN (8 c) (5.2-1) 104
WAVELENGTH ffl) 22850 22800 22750
Epsilon Tau K1 III
00 CO
4370 WAVENUMBER ( c m '1)
Figure 24. Kitt Peak FTS spectrum of e Tau in the region of the 2.101 pm bandpass of the infrared colors program. The top spectrum is a solar spectrum, the middle spectrum is the stellar spectrum, and the lower spectrum is the ratio of the two. Line positions are from the Hall (1970) atlas. ______RLPHfl ORI I. 0 14755' 4760' 14765' U77b' HRVENUMBER
Figure 25. Comparison of an observed spectrum of a Ori (solid line) with a synthetic spectrum (+ symbols) in the region of the 2.101 pm bandpass of the infrared colors program. The region scanned was from approximately 4751 cm""-*- (2.104 pm) to 4765 cm“l- (2.098 pm). Telluric absorption features are indicated below the spectrum. The syn thetic spectrum was computed with CN lines only. 106
Table 12
Comparison of 2.1 pm Bandpass Depressions and 8-color CN Indices
2.101 pm Bandpass 8-color Depression CN Index Star (mag) (mag) a Tau 0.032 0.086 p Gem 0.033 0.072 a Sco 0.069 0.190 a Ori 0.072 0.178 a Her O.O69 __ 107
where the 8 -color CN index is defined to be the mean of the depressions observed at filters 4 and 8 in magnitudes (White and Wing, 1978).
2.2850 pm (4375 cm-1). This bandpass is just.shortward of the
(2,0) bandhead of 12C0 at 4360 cm-1. This region in e Tau is shown in
Figure 26. Although a moderate amount of telluric absorption occurs in this region, only a few weak stellar features are observable at high resolution. Magnitudes reported for this wavelength are the average of scanner data obtained in second order at five points between 2.282 pm
(4381 cm-1) and 2.288 pm (4370 cm-1). They are designated I<(228) and o have an effective bandpass of about 80 A.
3.996 pm (2502 cm-1) . Continuum magnitudes at this wavelength, referred to as L(400), were obtained from the 4 pm scans of the SiO first-overtone bands. This spectral region is discussed in Section 4.2.
Section 5.3 - Observational Data
A total of 111 bright late-type stars were observed at one or more wavelengths of the infrared colors program. Measurements at 3.996 pm were made as part of the SiO continuous scan program described in
Chapter IV and at 1.287, 2.101, and 2.285 pm during 5 nights and 2 days in November 1977 with the infrared scanner. Data were obtained at all four infrared wavelengths for 50 stars.
The spectral types of the "oxygen-rich" program stars range from
G1 to M 8 with the majority being IC and M giants. Observations on the
Wing 8 -color system have been used to provide 1(104) magnitudes and a near-infrared color temperature. The spectral region covered by the 108
WRVELENGTH (R) 21000 20950
Epsilon Tau K1 III
o coo © oo © o o tip © © © co i I I s ] R ICR
_i__l___ .___ I .___ . I l_ j .___ .___ I--- .--- 1--- .--- .--- 1--- 1--- 1--- 1--- 1--- 1--- 1--- 1--- 1--- 1--- 1--- L H7BD M770 4780 4790 WRVENUMBER
Figure 26. Kitt Peak FTS spectrum of e Tau in the region of the 2.285 pm bandpass of the infrared colors program. The format is the same as Fig. 24. The 1 CO (2,0) bandhead can be seen at 4360 cm-^. 109
8 -color and infrared scanner data extends from 0.75 to 4.0 pm and includes the region where the majority of cool-star flux is emitted.
The observational data for non-Miras are shown in Table 13. The observed fluxes are expressed on a magnitude scale, and the zero points have been established by arbitrarily setting the magnitude of Vega equal to zero at each wavelength. Infrared color temperatures are listed below the magnitudes at each wavelength and have been computed from a blackbody fit to the fluxes at each wavelength and the reference wavelength 1.040 pm assuming the model atmosphere absolute colors of
Vega listed in Table 3. The color temperatures derived from the 8 -color data are listed in the second to the last column. The K(210) magni tudes have been corrected for CN contamination as described in Section
5.2.
The bandpasses for both the infrared measurements and the 8 -color continuum points have been carefully chosen to avoid as well as possible the effects of line blanketing. Despite this, the observed color tem perature derived between any two wavelength can differ from the "true" stellar continuum color temperature for the following five reasons: errors in the observed color index caused by (1 ) photometric errors; or (2) stellar variability; (3) systematic errors in the calibration of the absolute color of the primary standard star Vega; (4) differential line blanketing within the bandpasses; or (5) interstellar or circum- stellar reddening. These problems will be discussed next.
The photometric error in the 8 -color data should be very small
(<0.01 mag) since most stars have been observed repeatedly: in fact, Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J(129) K(210) K(228) L(400) T (8c) Notes
1013 45 X Peg M2 III 2.101 1.420 0.517 0.530 3623 3025 2921 3084
3627 165 6 And K3 III 1.551 1.012 0. 387 0.408 0.340 4270 3593 3566 3779 4260
3712 168 a Cas K0- Ilia 0.721 0.228 -0.326 -0.263 -0.342 4574 3843 3807 4139 4593
4128 188 B Cet K1 III 0.632 0.190 -0.290 -0.252 4799 4159 4109 4408
6860 337 3 And MO Ilia -0.405 -1.046 -1.963 -1.928 -2.033 3645 3166 2954 314 8 3557
9927 464 51 And K3 III 1.868 0.716 0.769 4297 3588 3871
11961 M5 III 1.931 1.168 0.103 0.083 -0.090 2623 2775 2651 2766 3095
12929 617 a Ari K2 Illab 0.416 -0.083 -0.693 -0.674 -0.748 4435 3810 3673 3891 4359 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras.
(cont'd.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J(129) K(210) K(228) L(400) T (8c) Notes
14270 AD Per M2.5 lab 3.974 1.744 2802 h&x Per 2900
14469 SU Per M3-4 lab 3.394 1.103 2708 h&x Per 2848
14488 RS Per M4.5 lab 3.648 1.270 2482 h&x Per 2777
14528 S Per M4e la 3.254 0 .010 2141 h&x Per 2238
17506 834 T] Per K3- Ib-IIa 1.416 0.788 -0.053 -0.003 0.003 3589 3215 3071 3296 3884
18191 867 45 Ari M6- III: 0.747 -0.031 -1 .110 -1.128 -1.270 2606 2733 2623 2739 3099
18884 911 a Cet Ml.5 III- -0.195 -0.854 -1.768 -1.762 -1.884 3570 3101 2935 3089 3476 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont'd.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J(129) K(210) K(228) L(400) Tc (8c) Notes
19058 921 p Per M4 Ilb-IIIa -0.296 -1.043 -2.045 -2.053 -2.213 3323 2820 2733 2862 3204
20797 1009 M0+ Ila 2.010 1.420 0.399 0.409 0.246 3354 2886 3043 3378
22049 1084 e Eri K2 V 2.478 2.089 1.665 1.657 5081 4557 4422 4599
22649 1105 S5,3 1.754 1.212 0.192 0.209 -0.071 3319 3579 2949 3118 3309
23475 1155 M2+ Ilab 1.051 -0.867 3198 3204
25025 1231 Y Eri M0 III 0.517 -0.150 -1.032 -1.020 -1.125 3690 3073 2965 3129 3538
27371 1346 Y Tau K0 IIlab 2.331 1.911 1.433 1.455 4897 4315 4174 4431 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont'd.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J (129) K(210) K(228) L (400)
28305 1409 e Tau K0 III 2.198 1.767 1.296 1.327 4865 4235 4163 4446
29139 1457 a Tau K5 III -1.455 -2.065 -2.903 -2.898 -3.020 3737 3285 3099 3260 3647
30959 1556 o 1 Ori M3S 1.227 0.476 -0.573 -0.585 -0.787 3089 2806 2680 2803 3103
32068 1612 £ Aur K5 II+B 1.532 0.899 0.029 0.083 -0.039 3688 3196 3025 3252 3638
GP Ori SC 4.439 1.699 2247 2520
32736 1648 W Ori C5,3 2.028 -0.927 2889 2390
36389 1845 119 Tau M2 Iab-Ib 0.823 0.112 -0.991 -0.996 -1.228 2904 2928 2666 2796 3065 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont* d.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J(129) K(210) K(228) L(400) Tc (8c) Notes
38944 2011 U Aur M1+ Ilia 2.111 0.440 3576 3501
39801 2061 a Ori Ml-M2 Ia-Ib -2.679 -4.519 3235 3292
40239 2091 tt Aur M3 II 0.840 0.191 -0.834 -0.835 -1.069 3213 3137 2815 2955 3213
42475 2190 TV Gem M0-1 lab 2.919 2 .120 0.947 0.955 0.569 2673 2678 2516 2653 2799
42543 2197 BU Gem Ml-2 Ia-Iab 2.657 1.889 0.770 0.805 0.482 2943 2761 2594 2762 2948
42995 2216 ri Gem M3 III 0.156 -0.521 -1.500 -1.486 -1.654 3392 3039 2836 2995 3325
44478 2286 U Gem M3 Illab -0.273 -1 .000 -1.983 -1.972 -2.106 3385 2879 2775 2927 3298 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont'd.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD KR=BS Name Sp.Type 1(104) J (129) K(210) K (228) L(400) T (8c) Notes c
44537 2289 ip1 Aur K5-M0 Iab-Ib 2.046 0.250 3350 3343
48329 2473 e Gem G8 lb 1.256 0.737 0.091 0.151 4311 3702 3564 3858
55383 2717 BQ Gem M4 III 1.517 0.765 -0.246 -0.272 -0.383 3205 2806 2718 2828 3223
58061 VY CMa M5 I 2.771 -3.172 2095 1422
60414 2902 KQ Pup M2 Iabep+B 1.727 0.989 -0.060 -0.062 -0.254 3168 2846 2693 2828 3136
60522 2905 U Gem M0 III 1.680 1.039 0.188 0.225 0.103 3757 3166 3039 3243 3629
62509 2990 8 Gem K0 Illb -0.217 -0.663 -1.145 -1.108 -1.174 4784 4133 4092 4388 4870 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont'd.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J(129) K(210) K(228) L(400) T (8 c) Notes
69267 3249 B Cnc K4 III 1.463 0.863 0.101 0.106 -0.005 3925 3326 3225 3392 3794
76827 3576 p UMa M3 Illb 1.745 0.217 3478 3703
78712 3639 RS Cnc M 6e(S) 0.216 -1.979 2404 2931
80493 3705 a Lyn K7 Illab 0.834 0.178 -0.671 -0.655 -0.772 3772 3111 3022 3194 3588
81797 3748 a Hyd K3 II-III 0.010 -0.563 -1.322 -1.284 4005 3442 3273 3497
82308 3773 X Leo K5 III 2.016 1.342 0.519 0.549 3816 3049 3033 3226
84441 3873 £ Leo G1 II 1.867 2.195 1.159 1.167 5307 4923 4781 5026 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont'd.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J(129) K(210) K(228) L(400) Tc (8 c) Notes
85503 3905 U Leo K1.5 III 2.263 1.757 1.186 1.216 1.146 4419 3771 3740 3987 4464
86663 3950 it Leo M2- Illab 1.990 1.307 0.407 0.413 0.345 3574 3018 2922 3076 3534
90432 4094 U Hyd K5 III 1.696 1.092 0.295 0.311 3878 3310 3168 3348
95689 4301 a UMa K0- Ilia 0.317 -0.132 -0.719 -0.693 4517 4113 3831 4075
96833 4335 UMa K1 III 1.442 0.974 0.387 0.435 4514 3991 3789 4082
97778 4362 72 Leo M3 III 1.562 0.858 -0.106 -0.114 3421 2950 2822 2954
100029 4434 X Dra M0 III 1.345 0.669 -0.238 -0.215 3645 3042 2922 3098 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont’d.)
MK Magnitudes and Blackbody Temperatures (K) ID or BD HR=BS Name Sp.Type 1(104) J (129) K(210) K(228) L(400) T (8 c) Notes c
108849 BK Vir M7- 1.060 -1.298 1922 2793
112300 4910 6 Vir M3 III 0.313 -1.306 -1.302 3437 High 2879 3028 vel. ?
113226 4932 e Vir G8 Illab 1.602 1.229 0.804 0.833 5009 4695 4468 4772
119228 5154 83 UMa M2 Illab 1.898 1.284 0.322 0.350 3539 3270 2931 3114
124897 5340 a Boo K2 nip -1.820 -2.382 -3.066 -3.048 -3.128 4175 3491 3416 3614 4069
126327 RX Boo M 8 0.267 -0.859 -2.031 -2.093 -2.407 1631 2043 2259 2333 2563
127665 5429 p Boo K3 III 1.808 1.240 0.527 0.566 4215
3464 3355 3588 118 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont'd.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J (129) K(210) K(228) L(400) T (8 c) Notes c
131873 5563 6 UMi K4 III 0.028 -0.638 -1.425 -1.387 3944 3076 3093 3302
132813 5589 RR UMi M5 III 0.735 -0.053 -1.066 -1.070 3140 2707 2678 2811
146051 6056 6 Oph M0.5 III 0.221 -1.423 3678 3537
148387 6132 U Dra G8 III 1.486 0.587 0.589 4913 4170 4371
148783 6146 g Her M 6- III -0.191 -0.990 -2.088 -2.107 2495 2678 2584 2699
148856 6148 8 Her G8 III 1.493 1.075 0.628 0.697 5015 4329 4265 4680
156014 6406 a 1 Her M5 Ib-II -1.668 -3.785 2565 3003 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont' d.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J(129) K(210) K(228) L(400) Tc (8 c) Notes
156283 6418 it Her K3 Ilab 1.226 0.638 -0.095 -0.058 4038 3376 3291 3514
154143 6337 M3- III 2.055 1.352 0.431 0.436 3504 2953 2872 3023
159181 6536 6 Dra G2 Ib-IIa 1.507 1.210 0.748 0.761 4969 5051 4598 4853
161096 6603 6 Oph K2 III 1.213 0.675 0.125 0.168 0.126 4524 3605 3718 3992 4535
164058 6705 y Dra K5 III 0.042 -0.600 -1.413 -1.371 3858 3162 3090 3305
165674 VX Sgr M4e 2.302 -1.842 1592 1872
167006 6815 104 Her M3 III 1.981 1.231 0.277 0.279 3451 2811 2781 2924 120 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont'd.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J (129) K(210) K(228) L(400) Tc (8 c) Notes
172167 7001 a Lyr A0 V 0.009 0.000 0.000 0.000 -0.003 8892 20102 12035 12356 1 2100
172380 7009 XY Lyr M4-5 II 1.560 0.744 -0.346 -0.369 -0.505 2792 2635 2576 2686 3052
175588 7139 6 2 Lyr M4 II 0.502 -0.259 -1.314 -1.316 3181 2780 2663 2797
175865 7157 R Lyr M5 III -0.323 -1.115 -2.165 -2.175 -2.319 2951 2696 2638 2762 3121
186791 7525 Y Aql K3 II 0.675 0.078 -0.703 -0.679 -0.759 3927 3339 3201 3397 3849
+39°4208 RW Cyg M3-4 Ia-Iab 2.981 1.906 0.499 0.454 -0.021 2069 2121 2136 2222 2364
196610 7886 EU Del M 6 III L-l 0.706 -0.124 -1.192 -1.225 -1.393 2462
2600 2584 2684 3019 121 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont1d.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J(129) K(210) K(228) L(400) T (8 c) Notes
197989 7949 e Cyg K0- III 1.061 0.577 0.002 0.058 4640 3895 3779 4092
IRC+40448 M 6 7.189 -1.874 1015 1011
206778 . 8308 E Peg K2 lb 0.475 -0.183 -0.938 -0.884 -0.977 4098 3104 3149 3389 3820
206936 8316 U Cep M2 la -0.105 -0.924 -1.982 -1.953 -2.292 2747 2627 2603 2766 2938
207076 M7 III: 0.349 -0.559 -1.628 -1.687 -1.907 2158 2425 2512 2588 2877
208816 8383 W Cep M2 la-lab 1.591 0.866 -0.212 -0.199 -0.435 3017 2885 2676 2826 3090
210745 8465 C Cep K1.5 lb 1.319 0.730 -0.034 0.032 3879 3372 3240 3509 122 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont'd.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J(129) K(210) K(228) L(400) Tc (8c) Notes
212466 RW Cep K0 0-Ia 3.214 2.749 1.887 1.923 1.461 3090 4010 3280 3502 3394
213310 8572 5 Lac MO Iab+B 1.662 1.025 0.052 0.094 3484 3181 2890 3088
214665 8621 M4 III 1.493 0.826 -0.224 -0.192 -0.412 3287 3073 2767 2944 3217
216946 8726 MO- lb 2.321 1.702 0.749 0.793 0.584 3456 3250 2937 3141 3414
217476 8752 GO la 2.585 1.126 3667 3810
217906 8775 3 Peg M2.5 II-lil -0.592 -1.349 -2.338 -2.340 -2.443 3408 2791 2736 2872 3278
221615 8940 71 Peg M5- Ilia 1.502 0.767 -0.284 -0.273 -0.427 3170 2855 2694 2843 3191 123 Table 13
Narrow-Band Magnitudes and Color Temperatures for Non-Miras
(cont'd.)
MK Magnitudes and Blackbody Temperatures (K) HD or BD HR=BS Name Sp.Type 1(104) J(129) K(210) K(228) L(400) Tc (8c) Notes
224427 9064 Y Peg M3 III 1.629 0.900 -0.071 -0.062 3494 2873 2786 2937 125
many are 8-color standards. However, some of the cooler stars, parti cularly the supergiants, are slightly variable in the near infrared.
For many of these stars simultaneous 8-color observations were made in
November 1977 with the 0.4-m Kitt Peak telescope. The simultaneous data have been used, when available, so that the observational results refer to a single epoch. Errors may be somewhat larger for the cooler stars for which simultaneous 8-color data could not be obtained. The probable errors of the infrared magnitudes are about 0,016 mag
(Table 5). Unfortunately, because of the limited amount of observing time available, most stars were observed only once.
The errors in the absolute colors of the fundamental standard star
Vega are difficult to estimate. Since the absolute flux of Vega has been measured in the near-infrared (Hayes and Latham, 1975 ; Tiig, White, and Lockwood, 1977), the near-infrared colors are probably accurate to
0.02 mag. The situation at wavelengths longer than 1.1 pm is less satisfactory since the absolute energy distribution of Vega has not been measured and can only be inferred from model atmosphere computa tions. Despite the surprisingly large difference in the parameters deduced for Vega by Kurucz (1979) and Dreiling and Bell (1980), the infrared colors computed for their models are in good agreement and closely match the results of Schild, Peterson, and Oke (1971). It is therefore likely, but by no means certain, that the infrared absolute colors of Vega are accurate to ±0.03 mag. Additional discussion of this problem was presented in Section 2.4. 126
The occurrence of line blocking within the bandpasses used to com pute the color indices will cause the observed color temperature to differ from the continuum color temperature. As noted previously, the
color temperatures on the Wing 8-color system are now believed to be
systematically lower than the effective temperatures for all M stars
because of some source of blanketing in the 0.7540 pm filter (Ridgway,
et al. , 1980). The bandpasses of the infrared points have been examined
at high resolution (Section 5.2) and are believed to be nearly free of
line blocking in all but the coolest stars. Small corrections for CN
contamination have been applied at 2.101 pm. For giants and supergiants
later than about M6 it is important to remember that water vapor ab
sorption occurs throughout the infrared and is likely to effect one or more of the program bandpasses.
Table 14 has been constructed to show the effect of a given error
in the photometry, arising from any cause, on the resulting color tem
perature. The value 0.02 mag has been chosen as representative of
errors in the relative photometry. The error AT in the computed black-
body temperature is tabulated for the two shortward continuum points
of the Wing 8-color system (0.7540 and 0.7810 pm) and the four wave
lengths of the infrared colors program. As in Table 13 the reference wavelength for fitting the blackbody curve has been taken to be 1.040
pm. For relatively small errors in the color (<0.10 mag), the error
in the calculated color temperature is nearly linearly proportional
to the error in the color. As noted by Wing (1967a), the error A0
(6=5040/T) is nearly independent of temperature at each wavelength. 127 Table 14
Error in Temperature Resulting from a 0.02 mag Error in the Color Index taken with respect to 10395 X
Wavelength AT for T (A) A0 2000 2500 3000 3500 4000 4500 5000
7540 0.018 14 22 32 45 59 77 88
7810 0.020 16 25 37 51 68 88 95
12870 0.038 28 45 66 92 124 162 140
21010 0.015 11 18 26 37 50 65 81
22850 0.014 10 16 24 34 46 61 77
39965 0.011 8 13 19 27 36 47 59 128
The value of A0 for each wavelength is also listed in Table 14. Errors in the color temperature are largest at 1.287 pm because of the rela tively short baseline used to fit the blackbody curve.
The effect of reddening on the observed color temperatures has been calculated from the van de Hulst theoretical reddening curve #15
(Johnson, 1965). The absorption in magnitudes at each of the program wavelengths is related to the color excess in B-V by the relations
A (0.7540 pm) = 1.84 E(B-V)
A (0.7810 pm) = 1.72 E(B-V)
A (1.0395 pm) = 1.05 E(B-V)
A (1.287 pm) = 0.75 E(B-V)
A (2.101 pm) = 0.31 E(B-V)
A (2.285 pm) = 0.27 E(B-V)
A (3.9965 pm) = 0.12 E(B-V).
In Table 15 the change AT in the color temperature has been eva luated at the program wavelengths for color excesses in E(B-V) from
0.01 to 0.50. As expected from the wavelength dependence of inter
stellar extinction and the Planck functions the largest changes occur at the shortest wavelength and for the warmest stars. Results from this table have been used to compute the slope of the reddening line in the color temperature vs. color temperatue plots presented in
Section 5.5. 129
Table 15
Effect of Reddening on Color Temperatures
Wavelength AT for T E(B-V) (A) 2000 2500 3000 3400 4000 4500 500(
0.01 7540 6 9 13 17 23 30 38 7810 5 8 12 16 23 29 37 12870 4 7 10 13 19 24 31 21010 4 6 10 13 18 24 30 22850 4 6 9 13 18 23 30 39965 4 6 9 11 17 22 27
0.05 7540 27 43 62 81 114 146 184 7810 27 42 61 79 111 143 179 12870 21 33 49 64 91 117 149 21010 20 32 47 62 90 116 147 22850 20 31 47 62 88 115 145 39965 18 29 43 57 81 106 134
0.10 7540 54 84 122 158 221 283 353 7810 52 82 119 153 215 276 345 12870 41 65 95 125 177 228 288 21010 39 63 93 122 175 226 285 22850 39 62 92 121 172 223 281 33965 35 57 85 111 159 206 260
0.20 7540 105 163 234 301 417 529 656 7810 102 159 228 293 407 517 641 12870 81 127 185 240 337 433 541 21010 77 123 180 235 333 428 536 22850 76 121 178 232 329 423 529 39965 69 111 164 215 305 393 492
0.30 7540 154 237 338 431 592 747 918 7810 150 231 329 420 579 730 898 128 70 119 185 268 346 484 616 765 21010 114 179 262 340 478 610 758 22850 112 177 258 335 471 602 749 33965 102 163 239 311 439 561 699
0.40 7540 200 306 433 550 751 940 1148 7810 195 298 423 537 734 920 1125 12870 155 241 346 445 618 782 965 21010 148 233 338 437 610 774 957 22850 146 230 334 432 602 765 946 33965 133 212 310 402 562 715 886 130
Table 15
Effect of Reddening on Color Temperatures
(cont'd.)
Wavelength (A) 2000 2500 3000 3400 4000 4500 5000
7540 244 371 522 660 894 1113 1352 7810 237 362 510 645 875 1091 1326 12870 190 294 420 538 741 933 1145 21010 182 284 410 528 731 924 1136 22850 179 280 405 521 723 913 1124 131
Section 5.4 - Computation of Synthetic Colors
Since the continuum points were chosen in spectral regions nearly free of line blanketing, synthetic color temperatures and indices have been computed with only continuum opacity sources. The sources of con tinuous opacity included were H bound-free, H- bound-free and free-free,
H£- free-free, H Rayleigh scattering, H2 Rayleigh scattering, He Ray leigh scattering, He- free-free, H2+ bound-free and free-free, and electron scattering. The algorithm to compute the absorption and scat
tering coefficients was taken almost entirely from the synthetic spec
trum program MOOG (Sneden, 1974). The calculated H2- coefficients of
Somerville (1964) have been increased by 40% (John, 1975). Rayleigh
scattering by H2 was approximated by the formula of Dalgarno (1962).
The model atmospheres of Bell, et al. (1976a) and Johnson, Bernat, and Krupp (1980) were used to compute the emergent fluxes. In both grids the standard assumptions of plane-parallel stratification in homogeneous stationary layers in hydrostatic and local thermodynamic equilibrium were made. In the models of Bell, et al. (1976a) the line opacity was treated with opacity distribution functions whereas
Johnson, Bernat, and Krupp (1980) have used the opacity sampling method.
These models are believed to be the best currently available. Synthetic colors were also computed with several straight mean opacity models of
Johnson (1974), but the calculated colors were found to be in poor agreement with the observational data.
Models in the effective temperature range 2750 to 5000 K were con
sidered. When the appropriate models were available, synthetic colors 132
were computed for a range of gravities and chemical compositions at each effective temperature. The calculated synthetic color temperatures and color indices are tabulated in Table 16. Both the indices and color temperatures have been computed with 1(104) as the reference wavelength.
The color temperatures were computed by fitting a blackbody curve to the model fluxes between each wavelength and the 1.04 pm reference wave length. The color indices are magnitude differences expressed on a scale of absolute flux per unit wavelength.
To illustrate the wavelength dependence of the continua in late- type stars, continuum fluxes have been computed between 0.6 and 4.0 pm with the solar composition T^ = 3750 K and log g = 2.25 model of Bell, et al. (1976a). In Figure 27 the emergent fluxes (dots) are expressed on a magnitude scale per unit wavelength, normalized to the 1.0395 pm reference wavelength. Arrows mark the wavelengths of the infrared colors program and the continuum points of the 8-color system. For comparison, a 3750 K blackbody energy distribution (solid line) normalized to 1.0395 pm is also shown.
A peak in the continuous flux near 1.65 pm is clearly visible in the calculated continuum. The shape of the energy distribution can be readily explained by the wavelength dependence of the dominant infrared opacity source, H- (Figure 28). Absorption by H- at wavelengths less than 1.65 pm is primarily by the bound-free component. The bound-free cross-section is largest at 0.85 pm and falls to zero at wavelengths longer than 1.65 pm (Geltman, 1962). Free-free absorption occurs at all wavelengths: its cross-section increases to longer wavelengths but Table 16
Synthetic Colors and Blackbody Temperatures
Blackbody Temperature (K) and Color ■ Index (mag) (reference wavelength 10395 A) T log g Model 7540 7810 12870 21010 22850 33965 A eff
5000 1.5 Bell (DBV=2) 4909 4863 4112 4266 4401 4994 Solar -0.540 -0.489 0.419 1.840 2.165 4.431
2.25 Bell (DBV=2) 4898 4852 4067 4214 4357 4966 Solar -0.537 -0.487 0.412 1.822 2.150 4.421
3.0 Bell (DBV=2) 4883 4837 4037 4178 4326 4939 Solar -0.534 -0.484 0.407 1.809 2.139 4.412
4500 0.75 Bell (DBV=2) 4441 4406 3662 3752 3871 4401 Solar -0.428 -0.392 0.341 1.633 1.953 4.200
1.5 Bell (DBV=2) 4424 4385 3658 3698 3828 4371 Solar -0.423 -0.387 0.340 1.607 1.933 4.186
2.25 Bell (DBV=2) 4406 4366 3624 3653 3788 4341 Solar -0.418 -0.382 0.334 1.586 1.913 4.172
2.25 Bell (DBV=2) 4347 4310 3607 3620 3751 4290 A/H=-l.0 -0.402 -0.369 0.330 1.569 1.895 4.149
2.25 Bell (DBV=2) 4376 4339 3626 3647 3779 4324 |_i A/H=-l.0 -0.410 -0.376 0.334 1.583 1.909 4.164 £ Table 16
Synthetic Colors and Blackbody Temperatures
(cont'd.)
Blackbody Temperature (K) and Color Index (mag) (reference wavelength 10395 A) T log g Model 7540 7810 12870 21010 22850 33965 X eff
3.0 Bell (DBV=2) 4386 4346 3602 3622 3758 4314 Solar -0.413 -0.378 0.329 1.570 1.899 4.160
0.75 Bell (DBV=2) 3902 3868 3220 3163 3279 3756 Solar -0.261 -0.245 0.241 1.301 1.621 3.852
1.50 Bell (DBV=2) 3886 3852 3190 3112 3236 3725 Solar -0.256 -0.240 0.233 1.266 1.590 3.832
1.50 Bell (DBV=2) 3835 3805 3185 3097 3215 3699 A/H=-l.0 -0.237 -0.225 0.232 1.255 1.578 3.816
1.50 Bell (DBV=5) 3860 3829 3199 3117 3236 3725 A/H=-l.0 -0.246 -0.232 0.236 1.269 1.592 3.833
2.25 Bell (DBV=2) 3869 3834 3173 3083 3205 3701 Solar -0.250 -0.234 0.229 1.245 1.571 3.816
2.25 Bell (DBV=2) 3834 3802 3174 3080 3198 3681 A/H=-1.0 -0.237 -0.223 0.229 1.242 1.565 3.804
3.0 Bell (DBV=2) 3854 3819 3162 3065 3188 3684 Solar -0.244 -0.229 0.226 1.232 1.558 3.805 Table 16
Synthetic Colors and Blackbody Temperatures
(cont'd.)
Blackbody Temperature (K) and Color Index (mag) (reference x^avelength 10395 A) T log g Model 7540 7810 12870 21010 22850 39965 1 eff
0.0 Johnson 3641 3615 3051 2973 3075 3496 Solar -0.162 -0.159 0.196 1.162 1.474 3.672
1.0 Johnson 3647 3617 3025 2903 3015 3471 Solar -0.164 -0.160 0.187 1.106 1.426 3.652
2.0 Johnson 3653 3623 3036 2898 3013 3474 Solar -0.166 -0.162 0.191 1.102 1.425 3.655
0.75 Bell (DBV=2) 3611 3581 2977 2860 2970 3425 Solar -0.149 -0.146 0.173 1.070 1.390 3.616
1.5 Bell (DBV=2) 3601 3569 2960 2823 2937 3400 Solar -0.145 -0.142 0.167 1.038 1.361 3.596
1.5 Bell (DBV=2) 3562 3534 2941 2786 2904 3386 A/H=-l.0 -0.128 -0.129 0.161 1.006 1.332 3.586
2.25 Bell (DBV=2) 3589 3557 2918 2804 2918 3383 Solar -0.140 -0.138 0.164 1.021 1.345 3.582
2.25 Bell (DBV=2) 3574 3543 2952 2803 2917 3376 A/H=-0.5 -0.133 -0.132 0.165 1.021 1.344 3.578 Table 16
Synthetic Colors and Blackbody Temperatures
(cont'd .)
Blackbody Temperature (K) and Color Index (mag) (reference wavelength 10395 X) T log g Model 7540 7810 12870 21010 22850 39965 X eff
0.0 Johnson 3430 3407 2892 2762 2862 3272 Solar -0.069 -0.078 0.146 0.983 1.295 3.489
1.0 Johnson 3467 3441 2911 2742 2850 3284 Solar -0.086 -0.092 0.152 0.965 1.284 3.499
2.0 Johnson 3502 3476 2948 2773 2883 3319 Solar -0.102 -0.106 0.164 0.994 1.314 3.530
0.0 Johnson 3275 3257 2813 2630 2728 3120 Solar 0.007 -0.013 0.119 0.855 1.166 3.348
0.5 Johnson 3293 3272 2810 2611 2713 3118 Solar -0.002 -0.020 0.118 0.836 1.151 3.346
1.0 Johnson 3324 3302 2835 2629 2732 3143 Solar -0.018 -0.034 0.126 0.854 1.171 3.370
2.0 Johnson 3385 3363 2892 2689 2794 3205 Solar -0.048 -0.060 0.146 0.914 1.231 3.429 Table 16
Synthetic Colors and Blackbody Temperatures
(cont'd.)
Blackbody Temperature (K) and Color Index (mag) (reference wavelength 10395 $) T log g Model 7540 7810 12870 21010 22850 39965 X eff
3200 0.0 Johnson 3125 3110 2719 2504 2598 2970 Solar 0.088 0.057 0.085 0.719 1.028 3.194
1.0 Johnson 3218 3199 2789 2557 2657 3043 Solar 0.037 0.014 0.110 0.779 1.092 3.271
2.0 Johnson 3292 3272 2854 2632 2732 3119 Solar -0.002 -0.020 0.133 0.857 1.171 3.347
3000 0.0 Johnson 3033 3020 2682 2442 2533 2886 Solar 0.142 0.103 0.071 0.647 0.952 3.098
1.0 Johnson 3129 3113 2753 2506 2602 2964 Solar 0.086 0.055 0.097 0.722 1.031 3.187
2.0 Johnson 3213 3195 2822 2590 2686 3048 Solar 0.040 0.015 0.123 0.814 1.123 3.276
2750 1.0 Johnson 3041 3028 2718 2468 2557 2888 Solar 0.138 0.099 0.084 0.677 0.980 3.102 138
0.0
.0
2.0
3.0
4.0 8.0 2.0 3.0 4.0 WAVELENGTH (§Jm)
Figure 27. Comparison between the continuous energy distribution of the Bell e£ al. (1976a) solar composition T = 3750 K and log g 2.25 model atmosphere (dots) and a 3750 K blackbody energy distribution. The fluxes are on a magnitude scale per unit wavelength, normalized to the 1.0395 pm reference wave length. Arrows mark the wavelengths of the infrared colors program and the continuum points of the Wing 8-color sys tem. The peak in the continuum caused by the H- opacity minimum is clearly visible at 1.65 pm. 139
log k
1.0 2.0 3 .0 4 . 0 WAVELENGTH (pm)
Figure 28. Absorption coefficient of H“ (cm per neutral hydro gen atom) plotted against wavelength for a tempera ture of 3835 K and an electron pressure of 6.71x10“^ dynes/cm^. The solid line is the total H~ absorption coefficient; the dashed line indicates the free-free component only. The conditions correspond to a level in the Bell et_ al. (1976a) Tg = 3750 K and log g = 2.25 solar composition model close to continuum op tical depth unity at 1.0395 pm. 140
is less accurately known (Geltman, 1965; John, 1979). The sum of the two components has a sharp minimum at 1.65 pm. The occurrence of the
H- opacity minimum in combination with the location of the maximum of the Planck function at cool-star temperatures produces a sharp peak in the continuous flux at 1.65 pm. The "1.65 pm peak" can also be seen in the calculated continua of Tsuji (1978a). An examination of the calcu lated opacities indicates that H- is dominant in the deeper layers where the majority of the continuum radiation originates.
From an examination of Figure 27 it is apparent that color tempera tures computed with the 1.0395 pm reference wavelength will be strongly wavelength dependent. For all wavelengths longward of about 0.7 pm the continuum color temperature is less than the effective temperature; the largest difference between the color temperature and effective tempera ture occurs at the II- opacity minimum. For the model illustrated in
Figure 27 the color temperature between 1.04 and 1.65 pm is 1400 I< less than the effective temperature.
From Table 16 it can be seen that these same conclusions are valid for all the cool star models. The color temperatures are very wave length dependent, and the largest difference between the effective tem perature and color temperature occurs at the program wavelengths near the H- opacity minimum. The temperatures from the 8-color and 1(104)-
L(400) indices are predicted to be good indicators of the effective temperature.
Surface gravity appears to have only a minor effect on the calcu lated colors. For a given effective temperature a difference of less 141
than 100 K in the color temperature is calculated to occur between gravities believed to be typical of giant and supergiant stars. The differences in the computed color indices is calculated to be on the order of the photometric accuracy of the data. Also, the effect of gravity is even smaller when the calculated colors are plotted against each other. Models of various surface gravities and effective tempera tures define nearly a single curve on all color-color plots. This result can be seen in Figures 29 and 30. This is very helpful since the observed colors can then be compared to a single theoretical rela tion independent of surface gravity.
For the warmer temperatures (T > 3750 K) the grid of Bell, et al.
(1976a) was used to study the effect of differences in the chemical composition on the model colors. The shift in the synthetic colors with a change of (A/H) from 0.0 to -1.0 was found to be small
(typically 0.01 mag). As noted by Gustafsson, et al. (1975), the rather small sensitivity of the model structures to changes in (A/H) can be explained by saturation of spectral lines for near-solar compositions and the partial compensation for the increased line absorption by a corresponding increase in the continuous opacity. Although the chem ical compositions of late-type giants and supergiants are largely unknown, these calculations suggest that the infrared continuum colors should be only slightly affected by moderate differences in chemical composition, at least for near-solar-composition cases. Unfortunately, similar computations could not be carried out with models in the lower temperature range since only solar composition models were available. 142
The importance of the microturbulent velocity parameter was also examined. Continuum colors of models computed with a Doppler broaden ing velocity (DBV in Table 16) of 2 km/s and 5 km/s differed only slightly as can be seen in Table 16.
It is important to recall that the model atmospheres used to compute the synthetic colors make the standard assumptions of plane- parallel geometry, local thermodynamic equilibrium, hydrostatic equili brium, and horizontal homogeneity. Departures from these assumptions are likely to produce errors in the calculated colors, particularly for stars of the highest luminosities and coolest temperatures. These effects have been discussed elsewhere (cf. Gustafsson, et al., 1975;
Carbon, 1979) and will not be considered further here. Chromospheric free-free and bound-free emission (Humphreys, 1974; Gilman, 1974), thermal reradiation from circumstellar grains (Fawley, 1977), and/or
H2O thermal emission (Tsuji, 1978b) may also occur in the infrared, but these effects are likely to be important only for the most luminous stars.
Section 5.5 - Comparison between Observed and Synthetic Color Tempera
tures for Giant and Bright Giant Stars
In Figures 29 and 30 synthetic color temperatures generated with solar-composition model atmospheres are compared with the observed values for luminosity class II and III stars. Results obtained with the grid of Bell, et al. (1976a) are indicated with open symbols while filled symbols represent values obtained with the models of O ( S04-2I0) Figure 29. The relation between T (104-210) and T (104-400). Stars Stars (104-400). T and (104-210) T between relation The 29. Figure 2000 0 0 0 4 3000 I” 3000 models from the grid of Bell et_ al. (1976a)9 large solid solid large et_ (1976a)9 Bell of grid al. the from models T (104-400). The arrow indicates the approximate slope slope approximate the indicates arrow The = (104-210) T (104-400). relation T the by defined is line dashed The and dotSp respectively. Large open symbols represent represent symbols open Large respectively. dotSp sfiownsymbols and are + III as and 5l classes luminosity of symbols models computed by Johnson by computed models symbols Solid lines connect models with the same surface gravity. gravity. surface same the with models connect lines Solid o? the reddening line. reddening the o? (1980). All models were computed with solar abundances. abundances. solar with computed were models All (1980). 3000 3 0 0 0 EC 0 0 0 3 3 S 0 0 EC 0 0 S 3 Te 0 5 7 3 3 8 0 0 EC 0 0 8 3 C 104-400) EC 4000 „ Bernat, and Krupp Krupp and Bernat, 25 □ 5 .2 2 MODEL RESULTS Boll EC 0 0 5 4 Johnson SC 0 0 0 5 5000 143 t (104-400) 4000 - 4000 3000 - 3000 5000 - 5000 iue 30 Figure l_ o Terlto ewe c(0-0) n c(c frmdl ad tr of stars and formodels (8c) Tc and (104-400) Tc between relation The . h ahd ie s c(0-0) c(8c). Tc = (104-400) Tc is line dashed The uioiycass Iad I, hw i tesm fra sFg 29. Fig.as format same the in shown III, and II classes luminosity 2.25 □ 2.25 . O 3.0 . A 1.5 Bell MODEL logg + IT+ o HI RESULTS Johnson lo9 9 0.0 0.0 . 2.0 . B 1.0 & 3000 , 3000K 0 ,,0 0 0 3 / Tr (8 3600 K 3600 //& c) 3750 3750 K /3800 /3800 K 4000 4000 K 4000 / 4500 K 4500 5000 K 5000 —I I I— I—
5000 145
Johnson, Bernat, and Krupp (1980). Solid lines connect models with the same value of log g.
In the T (104-210) vs. 1^(104-400) diagram (Figure 29) the observed and synthetic relations are in close agreement for all temperatures.
The two color temperatures are nearly linearly related to each other with T (104-210) between 500 and 800 I< lower than T (104-400). This c c indicates that the model atmospheres can satisfactorily reproduce the observed "excess" flux near the H- opacity minimum and confirms that
H- is an important continuous opacity source in the cooler late-type stars.
The observed and model-atmosphere results in the 1^(104-400) vs.
Tc (8c) diagram (Figure 30) are in good agreement for the warmer stars but deviate systematically for the later spectral types. The devia tion begins in the early M stars and increases to cooler temperatures.
The calculations indicate that the two color temperatures should be nearly equal, whereas the observed relation turns sharply leftward for
Tc (8c) < 3300 K.
The results for the other diagrams are similar. Mien color tem peratures formed between 1.0395 urn and the infrared wavelengths are plotted against each other, the observed and model atmosphere relations agree to within about 100 K. Model atmosphere calculations based only on continuous opacity sources, however, cannot reproduce the rapid decrease in Tc (8c) relative to the other color temperatures observed in the coolest stars. 146
Mean relations between color, color temperature, MK spectral type, and effective temperature are given in Table 17 for luminosity class
III stars. The Ridgway, et al. (1980) calibration has been used to relate 8-color temperature to MK spectral type and effective temperature.
The relations between the color temperatures are well-defined except for T (104-129). In this case, the separation between the two wave lengths used to form the color temperature is small, and the photome tric errors have a large effect on the calculated color temperature, particularly for the warmer stars (see Table 14). At the other wave lengths the color temperatures are sufficiently well determined to
separate stars differing by one or two spectral subtypes in almost all cases. The few bright giants observed have color temperatures that are close to the mean giant-star relations.
An important check on the temperature scale of Ridgway, et_ al.
(1980) is to test if the calibration can be correctly reproduced by model atmosphere calculations. In Figures 31 to 35 the data of Table
17 are compared with results obtained with solar-composition model atmospheres. When the color temperatures formed between the 1.0395 pm reference wavelength and the infrared wavelengths are plotted against effective temperature, the observed mean giant-star relations (assuming
the Ridgway, et_ al. , 1980, calibration) agree with the model atmosphere
calculations to within about 100 K over the entire range of spectral
types. The surface gravity has only a minor effect on the model
atmosphere continuum color temperatures. Considering the uncertainties
involved, the agreement must be considered excellent. When a similar Table 17
Relations between Spectral Type, Color Index, and Temperature for Class III Giant Stars
Color Temperature (color index) Effective Sp. Type Tc(8c) (104-129) (104-210) (104-228) (104-400) Temperature
G8 III 4900 4310(0.42) 4190(0.89) 4510(0.85) 4930 K K0 III 4760 4110(0.45) 4020(0.96) 4330(0.91) 4830(0.97) 4790 K1 III 4580 3910(0.48) 3820(1.04) 4100(1.00) 4600(1.06) 4610 K2 III 4400 3720(0.52) 3640(1.13) 3890(1.09) 4400(1.15) 4450 K3 III 4200 3520(0.55) 3450(1.23) 3650(1.21) 4160(1.26) 4270 K4.0 III 4000 3350(0.59) 3260(1.34) 3440(1.33) 3920(1.39) 4095 K5.0 III 3860 3250(0.62) 3120(1.43) 3300(1.42) 3750(1.50) 3980 M0.0 III 3750 3150(0.64) 3030(1.50) 3210(1.48) 3640(1.57) 3895 Ml.0 III 3640 3060(0.67) 2950(1.56) 3120(1.54) 3540(1.64) 3810 M2.0 III 3530 2980(0.69) 2880(1.62) 3030(1.61) 3430(1.72) 3730 M3.0 III 3400 2880(0.73) 2790(1.70) 2930(1.70) 3300(1.83) 3640 M4.0 III 3250 2800(0.75) 2710(1.77) 2850(1.77) 3220(1.90) 3560 M5.0 III 3000 2770(0.76) 2630(1.85) 2780(1.83) 3130(1.99) 3420 M6.0 III 2600 2730(0.78) 2600(1.88) 2720(1.89) 3100(2.02) 3250 m iue 1 T 1419 potdaant fetv tmeaue The temperature. effective against plotted (104-129) T 31. Figure ( ( 104-129) 4000 3000 - 3000 00 5000 0 0 0 4 3000 where the color temperature is equal to the effective effective the to equal is temperature color the where atmosphere results (Johnson, Bernat, and Krupp 1980) are are 1980) Krupp and Bernat, (Johnson, results atmosphere are indicated. Opacity-sampling solar-composition model model solar-composition Opacity-sampling indicated. are andwt te oa-opsto oes f el et_ Bell of al. models ob solar-composition the values mark with tained symbols open symbols; solid by indicated (+ symbols) types spectral corresponding The relations III 17. class Table in luminosity the by defined is line solid temperature. (1976a). The dashed line defines the locus of points points of locus the defines line dashed The (1976a). 2.25 3.0 MODEL 1.5 Bell log FETV TMEAUE (K) TEMPERATUREEFFECTIVE g . a 1.0 o □ A RESULTS Johnson . o 2.0 0.0 0.0 / • ogg / a / A Jvl 2 148 o (104-210) iue 2 T 1420 v. fetv tmeaue s hw i the in shown is temperature effective vs. (104-210) Tc 32. Figure 0 0 0 3 0 0 0 4 00 0 0 0 4 3000 - - _ - - ae omt sFg 31. Fig. as format same .5 . H 1.0 □ 2.25 MODEL . o 3.0 . A 1.5 o glg g log g log Bell » M 6 FETV TEMPERATUREEFFECTIVE RESULTS Johnson . o 2.0 . A 0.0 —I —r i— I— i— — 1 / ' / i r i (K) 5000 149 h*
o (104 — 228) Figure 33. Tc (104-228) vs. effective temperature is shown in the in shown is temperature effective vs. (104-228) Tc 33. Figure 4000 3000 00 00 5000 4000 3000 ae omt s i. 31. Fig. as format same MODEL RESULTS FETV TMEAUE .( TEMPERATUREEFFECTIVE K) Johnson 150 iue 4 T 1440 v. fetv tmeaue s hw i the in shown is temperature effective vs. (104-400) Tc 34. Figure (104-400) 4000 5000 00 i 3000 00 4000 3000 ae omt s i. 31. Fig. as format same MODEL 4 .5 □ 2.25 3.0 3.0 Bell . A 1.5 i i i i i I i I i i i i ! •°g9 FETV TMEAUE (K) TEMPERATURE EFFECTIVE O RESULTS Johnson 0.0 0.0 . o 2.0 1.0a o g log a \ i l 5000 151 152
5000 — i— i— i— |— i— i— i— i— |— r A
- MODEL RESULTS - Bell Johnson Tr(8c) — !°g g !°g g - 1.5 A 0.0 A 2.25 □ 1.0 a 3.0 O 2.0 o 4000
3000
3000 4000 5000 EFFECTIVE TEMPERATURE (K)
Figure 35. Eight-color near-infrared color temperature vs. effective temperature is shown in the same format as Fig. 31. 153
plot is made with the 8-color data (Figure 35), agreement is excellent for the warmer stars (T > 3800 K ) , but again the models cannot repro duce the rapid drop in the 8-color temperature in the M stars. The dis crepancy between the observed and model atmosphere continuum color tem peratures is very large for the coolest stars (over 600 IC at T^ = 3250 K) .
This effect has also been noted by Ridgway, et al. (1980), who compared the 8-color temperatures to effective temperatures determined from angular diameters. The lunar occultation data led to an effective temperature scale in the M stars considerably warmer than the corres ponding 8-c.olor temperature scale. It was proposed that there exists a source of blanketing in the 0.7540 pm region that increases moderately from M0 to M 6 .
The present study provides additional evidence to support this hypothesis. The model atmospheres of Bell, et al. (1976a) and Johnson,
Bernat, and Krupp (1980) predict relations between the color tempera tures close to the observed values for all the infrared wavelengths but not for the 8-color system. The rapid decrease in 8-color temperatures relative to the infrared color temperatures in the M stars also suggests that an additional opacity source is affecting the 8-color results. In addition, the infrared color temperatures predicted with the model atmospheres are consistent with the effective temperature scale of
Ridgway, et al. (1980), whereas the computed 8-color temperatures are not. These discrepancies can only be resolved by assuming the 0.7540 pm region is affected by some source of blanketing. 154
The source of blanketing at 0.7540 uni is difficult to determine.
A continuous opacity source seems unlikely since it would have to be moderately strong at 0.7540 um but negligible at 1.040 pm, since the infrared color temperatures computed with the 1(104) magnitude appear to be unaffected. Absorption by CN would act to lower the color tem perature but CN decreases in strength with decreasing, temperature in the M stars. The V0 molecule depresses the flux within the 0.7540 um bandpass in stars later than about M 6 , but its strong temperature dependence assures that the affect of V0 in the early M stars should be negligible.
Model atmosphere calculations including TiO line-by-line as an opacity source have been carried out by A. Bernat, J. Piccirillo, and
H. R. Johnson at Indiana University (as mentioned by Ridgway, et_ al.,
1980). The results indicate that the 0.7540 um region is depressed by high-excitation lines of TiO. Although these lines have never been observed in the laboratory, their calculated intensities indicate these lines are the likely source of the 0.7540 um depression.
A few supergiants and Mira variables were also observed, but the observational data are difficult to interpret. The supergiant colors are affected by interstellar reddening, and it is difficult to infer anything concerning their intrinsic colors. Also, many supergiants are known to be variable at 1(104) and in 8-color spectral type (White and Wing, 1978), and therefore are likely to vary in the infrared. A study of Mira variable colors would require observations at all phases and interpretation with dynamical models. 155
Of the color temperatures measured in this study, 1^(104-400)
appears to be the best indicator of the effective temperature. The
relation between T (104-400) and effective temperature is nearly c linear, and the model atmosphere calculations indicate that T (104-400) c should deviate from Tg by less than 300 K over the entire range of
spectral types studied here. Since the opacities at 1.0395 and 3.9965
pm are nearly equal and of the same origin (predominately continuous
absorption by H-), 1^(104-400) should be insensitive to differences in
the stellar temperature structure caused by differences in composition.
Also, 1^(104-400) is less affected by reddening than temperature indi
cators measured in the visual part of the spectrum, such as B-V.
Of course, the other color temperatures can also be calibrated
for use as indicators of the effective temperature. The 8-color tem
peratures are quite sensitive and are close to the Ridgway, et al.
(1980) effective temperatures for the warmer stars. For cooler stars
of abnormal composition, the 8-color temperatures should be used with
caution since the uncertain origin of the 0.7540 pm opacity could make
the observed values sensitive to gas pressure, microturbulent velocity,
and/or chemical composition. Because of the large difference in con
tinuous opacity between 1.0395 pm and wavelengths near the H- opacity
minimum, the other infrared color temperatures should be regarded as
less desirable temperature indicators. These color temperatures will
be sensitive to differences in the atmospheric temperature structure.
They may, however, be useful for other purposes. As noted by Bell, 156
et al. (1976b) , this sensitivity may offer an observational method of studying the effects of convective energy transport on the temperature structure of stellar atmospheres.
Previous studies of the infrared colors of late-type stars have been made at lower resolution. The H-flux peak is a pronounced feature in the spectrum of a Tau obtained by Woolf, et al. (1964) with the balloon-borne Stratoscope II telescope and has been studied by Bahng
(1969), Walker (1969), Lee (1970), Bell, et al. (1976b), and Persson,
Aaronson, and Frogel (1977) with intermediate and wide-band photometry.
The interpretation of the intermediate and wide-band results is complicated by line blocking, particularly in the 1,5-1.8 pm (H band) window, where the H- flux peak is expected to be largest. As noted by
Bell, et al. (1976b), the effect of CO and CN lines in this region is of crucial importance in interpreting the luminosity dependence of the colors reported by some observers (cf. Lee, 1970), since both absorbers are known to increase in strength with higher luminosity. Inspection of Kitt Peak and M t . Palomar FTS data (Connes and Michel, 1974;
Ridgway and Hall, 1978) confirms that this region is affected by numerous lines of CN and CO. In addition, absorption by atomic lines and OH is also present. The line blocking in this region increases rapidly to later types and exhibits a positive luminosity effect. In the dwarf stars absorption by H2O will occur in spectral types as early as about M2 (Mould, 1978) and should effect the wide-band colors of these stars. 157
Bell, et al. (1976b) have shown that line blocking must be included in the calculation of synthetic intermediate-band colors. The results of this study show that these problems can be avoided by mea suring at continuum points with narrow-band photometry, greatly sim plifying the interpretation of the observational data. CHAPTER VI
LABORATORY STUDIES OF THE INFRARED SPECTRUM OF ACETYLENE
Section 6.1 - Introduction
Absorption bands of the molecule acetylene (C2H 2) have recently been detected in the infrared spectra of several cool carbon stars
(Ridgway, et al., 1976; Ridgway, Carbon, and Hall, 1978). This chapter presents a laboratory analysis of the bands reported in these stars including measurements of intensities of some lines in the stronger bands.
The laboratory data were taken in 1978 and 1979 with The Ohio State
University 10-m focal length Czerny-Turner vacuum grating spectrometer in collaboration with Dr. K. Narahari Rao of The Ohio State University
Department of Physics and Dr. A. Baldacci of the Institute of Organic
Chemistry in Venice, Italy. This project was part of a ongoing study of molecules of astrophysical interest in the laboratory of Dr. Rao.
Dr. Baldacci has studied the infrared spectrum of acetylene for a number of years, and references to his work on this molecule appear throughout this chapter. During the summer of 1979 and in January and February
1980 the author worked in Venice, Italy, with Dr. Baldacci to complete the analysis at 3 pm reported in this chapter.
Acetylene is an important molecule in the atmospheres of many cool carbon stars. High-resolution spectra obtained at Kitt Peak indicate
158 159
that for some stars it is an important contributor to the strong 3 pm depression (Ridgway, Carbon, and Hall, 1978). Since polyatomic forma tion is favored at lower temperatures and 3 pm is close to the flux peak, it can be expected that acetylene will be an important opacity source in the upper atmospheres of the cooler carbon stars and should be included in model atmosphere computations. Also, since the v 3 band of 12C 13CH2 has been detected, the acetylene bands can be used to derive a value of the 12C/13C ratio in some stars.
As emphasized in the recent review article on infrared spectro scopy of stars by Merrill and Ridgway (1979), there is an important need for laboratory data of many atomic and molecular species of astro- physical interest. This study was undertaken to provide some of the line data needed for the molecule acetylene.
Section 6.2 - Theory of Vibration-Rotation Spectra of Linear Molecules
In the first approximation the term values (cm-1) for the rota tional levels in a vibration-rotation band are given by
V = = G (v, 2,, J) + F (2,, J) (6 .2-1 ) he v where G(v,2,,J) and F^(2,,J) are the vibrational and rotational term values, respectively. The rotational term value is given by
F (2, J) = B fJ(J+l) - 22] - D [J(J+1) - 2,2 ]2 + H [J(J+1) - &2]3 V V V V
- Lv [J(J+l) - l2]h + (6 .2-2) 160
where £ is the vibrational angular momentum quantum number which arises
for degenerate vibrations (v^ and V 5 modes of acetylene). For each degenerate mode with vibrational quantum number v_^ the vibrational angular momentum quantum number can assume the values
£. = v ., v . , v . ,...,1 , 0 1 1 1 - 2 1-14
In linear molecules the vibrational states may be classified according to the value of the quantum number £ as indicated in Table 18 below.
Table 18. Classification of Vibrational States of Linear Molecules
£ = I Ai Vibrational Species
0 E+ or Z~
1 n 2 A 3 $
•• •• ••
The existence of £ values greater than zero leads to an energy splitting
called "£-type doubling". In II vibrational states the splitting is
significant, but in A states the effect is usually small except when a
I vibrational state lies nearby. Since the total angular momentum J
must be equal to or greater than the vibrational quantum number £,
rotational levels within a vibrational band are given by
J = £, £+1, £+2, £+3, .... 161
Sublevels of II, A, $, ... states are designated as e or f depending on the symmetry of rovibronic eigenstates upon inversion. In this study the following convention is adopted for labeling e and f components: an e component is + if J is even and is - if J is odd and an f compon ent is - if J is even and + if J is odd.
In the case of acetylene ^he individual vibrational states are
£i £ c £ labeled as viv2v3vt v5 or V1V2V3(vtvs) • The latter designation applies to states involved in vibrational £-type resonance or doubling.
If two states exist with the same designation, they are distinguished with the aid of Roman numerals I and II rather than by the values of
£4 and £5 since they are no longer good quantum numbers.
The interpretation of the vibrational energies of acetylene is complicated by the existence of £-type resonance and doubling effects in the degenerate modes. Pliva (1972a) has studied the degenerate modes of 12C2H2 and has obtained a set of molecular constants from the laboratory data of Pliva (1972b) and Palmer, Miclcelson, and Rao (1972).
Energy expressions correct to the fourth order of approximation were required, and solutions were obtained as eigenfunctions of the vibra tion-rotation energy matrix.
The total eigenfunction in zeroeth approximation can be written as a product of the electronic, vibrational, and rotational eigenfunc tions. Rotational levels of linear molecules are designated as having positive or negative parity depending on whether the total eigenfunc tion remains unchanged or changes sign for an inversion of all particles at the center of mass. If the electronic and vibrational eigenfunctions 162
remain unchanged, the symmetry character depends only on the rotational eigenfunctions and the even rotational levels are positive, the odd ones negative. This is the case for states of acetylene, while £ vibra tional levels have negative even rotational levels and positive odd rotational levels. In n, A, etc., vibrational levels both positive and negative components occur for each J, and the order alternates for successive rotational levels etc. or etc.).
For linear molecules with a center of symmetry (point group D^), the total wavefunction can either be symmetric or antisymmetric with respect to a simultaneous exchange of all pairs of identical nuclei.
All vibrational levels that are symmetric with respect to an inversion are subscripted g (gerade) while those that are antisymmetric are subscripted u (ungerade). The positive rotational levels are symmetric and the negative antisymmetric for "g" states. The reverse is true for rotational levels within a "u" subscripted vibrational state.
These designations do not apply to linear molecules of point group C ^ , such as 12C13CH2 .
The dipole selection rules for vibration-rotation transitions of linear molecules have been summarized in Table 19-. The symbol -e > indicates that the transition is allowed and -<— -j— ► means it is forbid den. In the case for which A£=0 but 5.^0, AJ=0 is allowed, but the corresponding (Q) branch is weak with the intensity decreasing rapidly to higher J. 163
Transitions for which AJ = J ’-J" = +1 are termed R branch transi tions, AJ = 0 transitions are Q branch transitions, and AJ = -1 are P branch transitions. The single primed values refer to the upper level and the double primed to the lower level.
If a linear molecule belongs to point group D , that is, if it has “h a center of symmetry, alternate rotational levels have different statis tical weight due to nuclear spin. The ratio of the statistical weights of the symmetric and antisymmetric levels depends on both the nuclear spin and the type of statistics (Bose or Fermi) obeyed by the consti tuent nuclei. Statistical weights applicable to the various forms of acetylene of point group are summarized in Table 20. Since the rotational levels within a vibrational band are alternately symmetric and antisymmetric, successive lines in each branch of a molecular band of a D , molecule show a characteristic "alternation of intensities", °°h This alternation is very helpful in deciding the J numbering of a band and in identifying the vibrational states involved in a transition.
The effect of nuclear spin statistics must also be included in computing the partition function (see Section 6,5).
For isotopic forms of acetylene not listed in Table 20 the distinc tion between symmetric and antisymmetric levels disappears and there is no difference in the nuclear spin statistical weights of the even and odd rotational levels. These molecules are of the COT^ point group and intensity alternation does not occur in the spectra. 164
Table 19. Infrared Selection Rules for Linear Molecules
1. A& = 0, ±1 Vibrational Levels
2. X X
3. g <— p U, g ■*-/-+ g, U <5-7^ U
4. AJ = 0, ±1 (J=0 J=0)
5. + *—► + •»/❖ +, - -
6. s o— o s , a
Table 20
Nuclear Spin Statistical Weights for Acetylene Isotopes of Point Group D . °°h
- Statistical Weight Factors
Resultant Symmetric Antisymmetric Molecule Statis tics Levels Levels
1 2 r h Fermi 1 3 2 2
13r h Bose 10 6 2 2
12C D Bose 6 3 2 2
13C D Fermi 15 21 2 2 165
Section 6.3 - Measurement of Line Intensities of Linear Molecules
Consider a beam of radiation of intensity 1 (v) incident on a homp- o geneous gas sample in thermal equilibrium. The one dimensional equa tion of radiative transfer for pure absorption can be written (Arnold,
Whiting, and Lyle, 1969)
dlpX>dx') = a (vH 1-exP ( ^ B(v,T) -1- a(v) I(v,x) exp (
-a(v) I(v,x) (6.3-1) where a(v) is the absorption coefficient per unit length per unit pres sure at wavenumber v, I ( v jX) is the specific intensity at distance x measured from the boundary of the gas sample, and B(v,T) is the Planck function of the gas at temperature T. The first term on the right side corresponds to spontaneous emission, the second is induced emission and the last is absorption.
For a slab of thickness L, the emergent intensity is given by
I(v,L) = B(v,T) [l-exp(-k(v)pL)]+ Iq (v) [exp(-k(v)pL)] (6.3-2) where
k(v) = a (v) [1-exp ^ J is the absorption coefficient (cm-1 atm-1) corrected for stimulated emission. In the case that B(v ,T)< T(v) = = exp(-k(v)pL). (6.3-3) •Lq l‘) 166 For room-temperature absorption experiments in the 3 pm region, Equa tion 6.3-3 applies since the incident intensity of the spectrometer beam is far larger than the magnitude of the Planck function. The line intensity or the integrated absorption coefficient is defined by the relation (Pugh and Rao, 1976) (6.3-4) For laboratory work the units in common use are cm 1 atm 1 at tempera ture T for k(v), atm for pressure, cm for path length, and cm-2 atm-1 at temperature T for the line intensity. Other units and conversion factors are given by Pugh and Rao (1976). The intensity of a rotational line will change with temperature in accordance with the Maxwell-Boltzmann distribution law. The line in tensity at temperature T can be converted to a reference temperature Tq from the expression (6.3-5) where E is the excitation energy of the lower level of the transition above the ground state and Q(T) is the total internal partition func tion at temperature T and is given by Q(T) 2 (f)(2J+1) exp[-E(v,J)/kT] (6.3-6) v J 167 where v represents the set of quantum numbers which define a vibrational state and cj> is the nuclear spin statistical weight (1 or 3 for 12C2H 2). The summation over all levels includes rotational levels split by £-type doubling. For molecules in which one atom is replaced by an isotopic atom, for example 12C33CH2, there is no difference in the statistical weights of even and odd levels and The line intensity (cm-2 atm-1 at temperature T) for a rotational level within a vibration-rotation band of a linear molecule is given (Penner, 1959) by 8tt3v N, o t v'i'J1 JJ,£ R sj (i) ' 3hcQ(T) 5J'£' J £ v£ J hcv 1-exp Tf~ (6.3-7) where is the total number of molecules per unit volume per unit JU' pressure, V is the wavenumber at line center. is the matrix J£ element of the dipole moment of the pure rotational transition, v 1 £1 J' R is the matrix element of the dipole moment of the vibrational v£J transition and gji^i the statistical weight of the upper state. Applying the proper summation rules and including the nuclear spin statistical weight factor, the following expressions can be derived from Penner (1959). 168 For A£ = 0 (parallel bands), J ' £1 (m2-£ 2) R and P branches, 3J'£' J£ [ml J' £' £2 (2J+1) U Q branch (6.3-8) 3J'£' J £ J(J+1) where the unprimed quantum numbers refer to the lower level of the transition and the primed quantum numbers refer to the upper level, In the above m = J+l for the R branch and m = -J for the P branch. For A£ = ±1 (perpendicular bands), J ' £' (J ± £+ 1) (J + £ + 2) U R branch, J£ 2 (J+l) (-J + £) (J + £ - 1) P branch. 2J (2J+1) (J + £) (J ± £ + 1) Q branch. (6.3-9) 2[J(J+l)] If the line strength is measured for an unresolved £-type doublet, the factors given in Equation 6.3-8 must be multiplied by 2. Because of the interaction of vibration and rotation, the matrix element of the vibrational transition is often written as: v ' £' J1 v ' £1 R R F(J) (6.3-10) v£ J v£ v ' £ 1 where R is referred to as the rotationless matrix element and v£ F(J) is a factor which accounts for the interaction of vibration and rotation (centrifugal stretching, Coriolis interaction, etc.). 169 Band strengths are frequently reported in the literature. In this analysis the band strength is defined as the sum of the line intensities of all lines in all branches in a band: Stot(T) = 5 3 SJ (T)* (6.3-11) band If degenerate levels are involved, the summation is taken to include the subbands split by Ji-type doubling. Equation 6.3-11 is convenient when comparing results with low resolution measurements. Unfortunately, under the experimental conditions used in this study, the line intensity cannot be directly computed with Equations 6.3-3 and 6.3-4. The Doppler half-width is approximately a factor of 10 smaller than the resolution of the spectrometer used in this study. Therefore, the observed line profile primarily reflects the instrumen tal profile function, and the absorption coefficient cannot be directly measured as a function of wavelength. It is therefore necessary to assume a knowledge of the broadening mechanism and use the method of equivalent widths. The measured equivalent width is defined by the relation ' Vo+n / (1 - T(v))dv. (6.3-12) V — T) 0 In the above v is the wavenumber at line center and the integration- o has been carried out to distance r| (cm-1) from line center where the absorption is negligibly small, From Equations 6.3-3 and 6.3-12 the 170 equivalent width can also be calculated over the same interval from the absorption coefficient with the relation W [1 - exp (-lc(v)pL) ] dv (6.3-13) c It has been shown theoretically (DePrima and Penner, 1955; Goody, 1964) that the equivalent width is independent of resolving power if Iq varies linearly with v, if the response function is symmetric, or if Iq is constant over the width of the response function. For most lines the dominant broadening mechanism is either colli- sional or Doppler broadening or a combination of the two. The usual approximation is to adopt the Voigt profile which is a combination of independent Lorentz and Doppler broadening mechanisms. It is given (Pugh and Rao, 1976) by (6.3-14) where M is the molecular mass, is Avogadro's number, and b^ and bp are the half-width at half height of the Lorentz and Doppler profiles, respectively. In the present analysis the numerical relations of Armstrong (1967) were used to evaluate the Voigt function. The compu ted values agree with the tabulated values of Hummer (1965) to eight decimal places. 171 It is generally assumed that collisional half-widths are linearly proportional to the pressure so that bL = Yop. (6.3-15) Although techniques have been developed to calculate theoretical colli sional half-widths (cf. Yamamoto, Tanaka, and Aoki, 1969), it is generally agreed that laboratory measurements are preferable because of inadequacies in the theory (Varanasi and Sarangi, 1974). It is important to note that the values of y q are a function of rotational quantum number and the temperature, and the values are different for each molecule and broadening gas. Under laboratory conditions it is not possible to carry out the integration of Equation 6.3-12 over all wavelengths. Instead the integrals in Equations 6.3-12 and 6.3-13 are carried out to a distance q (cm-1) from line center where the absorption is small and changing slowly over one spectral slit width. In this case it is necessary to apply corrections for experimental loss resulting from the slow decrease of the absorption in the wings of Doppler-Lorentz lines. The situation is illustrated in Figure 36. Here a line with line center at has a measured equivalent width (singly cross-hatched area) determined by integrating from v„ = v - n to v = v + n. The J b I o u o measured 100% transmittance level has been determined from data points near the ends of the integration interval and is less than the true 100% transmittance level because of residual absorption near the end points. The measured equivalent width and the true equivalent width 172 CORRECTION FOR EXPERIMENTAL LOSS 100% LxJ O z: < i- CO z < cr t- LU V (c m '1) W (TRUE) = [l- T ( 2 /g ) -AZ/+T(I/f)- W(MEAS) W(TRUE)=TRUE INTEGRATED ABSORPTION W(MEAS) = MEASURED INTEGRATED ABSORPTION Figure 36. Correction for experimental loss. The singly cross- hatched region indicates the area under the line profile used to measure the equivalent width. Be cause of residual line absorption near the ends of the integration interval, the 100% transmittance level will be underestimated. The additional area (doubly cross-hatched region) must be included to determine the true equivalent width. 173 are related by the expression (Fridovich, et_ a l ., 1980) Wt = [1-t(v£)]”Av + t (v £) Wm (6.3-16) The line strengths in the present analysis were determined by iteration. From a starting value of S a computer program determined W^ by integrating Equation 6.3-13 over the interval of measurement. Iterations continued until agreed with Wt within 1 part in 101*. For the conditions used in this study experimental loss corrections amounted to a change of only 0.5 to 2% in the calculated value of S. Measurements of line intensities are best made at high resolution and at low pressure. When the pressure is less than about 1 torr the line shape is almost a pure-Doppler contour which has a very sharp wing cutoff. The corrections for experimental loss are much smaller for a Doppler line than for a Voigt or Lorentz line. For a Lorentz line near the linear absorption region, integrations out to a total distance of 25 Lorentz half-widths (of the order of 2.5 cm-1 at atmos pheric pressure for many molecules) are required in order to keep corrections for experimental loss less than 10% or over a distance of 400 Lorentz half-widths to obtain corrections of 0.3% (Korb, Hunt, and Plyler, 1968). At these distances line overlap is likely to be a problem and any small error in the baseline level will cause serious errors in the measurements. Since is an experimentally determined parameter, an error in its assumed value will result in additional error in the measured intensity. In addition, there is 174 considerable evidence that the Lorentz profile is only an approxima tion to the collision-broadened line shape in the far wings (cf. Winters, Silverman, and Benedict, 1964), and this will introduce an additional error. The measurement of Doppler lines requires high resolution. In Figure 37 data from the tables of Jansson and Korb (1967) have been used to generate curves of growth for several values of a. For typical conditions in this study a = 0.05 and b^ = 0.004. From the figure it can be seen that for equivalent widths larger than about 0.01 cm-1, the curve of growth becomes strongly nonlinear. Small errors in the measurement of the equivalent width can cause large uncertainties in the derived line strength at large equivalent widths. It is therefore important to measure weak lines (W <0.01 cm-1) at m high resolution. The resolution of The Ohio State University 10-m focal length Czerny-Turner vacuum grating spectrometer is sufficient (0.03 cm’"1) to make such measurements. = 10.0 a = 0. 4 Figure 37. Curves of growth. The data were taken from the tables of 175 Jansson and Korb (1968). 176 Section 6.4 - Experimental Details All acetylene spectra reported in this thesis were recorded with the Ohio State University 10-m focal length C z e m y - T u m e r vacuum grating spectrometer in double-pass mode. The experimental details have been summarized in Table 21. The data recorded in the 3 pm region were obtained at both room temperature and at a temperature of 160 C. The elevated temperature was used to bring out "hot" bands which might be visible in stellar spectra. A heating blanket and control unit purchased from the Briscoe Manufacturing Company (Columbus, Ohio) were used for the high temperature runs. A total of 32 charts were recorded for the line position studies. The experimental conditions and regions covered by each chart are summarized in Table 22. Chart #5 has not been included since it was later determined to have been recorded in the wrong grating order. Line positions were measured relative to the (1,0) and (2,0) bands of carbon monoxide (Mantz, et al., 1975). A Michelson goniometer (Chen, 1975) eliminated most of the sources of systematic error, especially changes in the rate of the chart paper and grating drives. Most line positions were measured at least four times, and some were measured ten or more times. Line positions for the stronger unblended lines varied by no more than 0.005 cm” 1. The mean measured positions of all 3000 lines are tabulated in Appendix B. A 99.6% minimum purity acetylene sample supplied by the Matheson Company was used for the measurements. For the intensity calculations, the gas sample was assumed to be 100% pure with the distribution of the 12 12 isotopic varieties given by the terrestrial values of 97.81% C 177 Table 21 Experimental Details in Recording Acetylene Spectra (a.) Spectrometer and Accessories 10-m focal length Czerny-Turner vacuum grating spectrometer equipped with a 40 cm x 20 cm (16" x 8") Harrison-ruled echelle (Smith et_ a_l. 1978) Source — carbon rod furnace (Rao 1972) Spectral resolution — about 0.03 cm Scanning rate — 0.01 cm ^/min; Detector — liquid nitrogen cooled InSb; Signal-to-Noise ratio — 50:1 peak-t.o-peak with a 3 sec time constant. (b.) Absorption Cells For intensity measurements — Glass cells equipped with CaF^ windows. For position measurements — Glass cell of 1-m length with CaF^ windows. Steel cell of 1-m length and a heating jacket for elevated temperatures. (c.) Gas Pressure For intensity measurements — 0.20 to 5.49 torr measured with MKS Baratron gauges. For position measurements — 2.5 to 100 torr pressure. (d.) Temperature For intensity measurements cell temperatures were monitored with precision thermistor probes. (e.) Range of Equivalent Widths for Intensity Studies 0.003 to 0.010 cm ^ corresponding to near Doppler lines on the linear part of the curve of growth. 178 Table 22 Experimental Conditions for 3 pm Position Spectra 2 Region Pressure Calibration (cm~l) Order (mm Hg) Temperature-*- Standard 3198-3254 18 3.5 R Cl, C2 2 3201-3259 18 3.5 R Cl, C2 3 3251-3297 19 2.5 R C2 4 3297-3357 19 2.5 R Cl, C2 6 3290-3366 19 2.5 R Cl, C2 7 3350-3407 19 45.0 R C2 8 3347-3405 19 45.0 R C2 9 3160-3209 18 45.0 R C2 10 3163-3209 18 45.0 R C2 11 3353-3402 19 50.0 H C2 12 3297-3357 19 2.5 H C2 13 3292-3360 19 7.5 H C2 14 3254-3292 19 2.5 H C2 15 3251-3292 19 2.5 H C2 16 3198-3259 18 2.5 H Cl, C2 17 3201-3259 18 2.5 H Cl, C2 18 3146-3192 18 100.0 H C2 19 3139-3198 18 100.0 H C2 20 3135-3192 18 100.0 H C2 21 3363-3405 19 100.0 H C2 22 3366-3407 19 100.0 H C2 179 Table 22 Experimental Conditions for 3 pm Position Spectra (cont'd.) 2 Chart Region Pressure Calibration No. ( cm- ) Order (mm Hg) Temperature^ Standard 23 3317-3376 19 12.0 H C2 24 3305-3357 19 12.0 H C2 25 3353-3378 19 12.0 H C2 26 3183-3259 18 12.0 H Cl, C2 27 3177-3254 18 12.0 H Cl, C2 28 3251-3305 19 7.0 Ii C2 29 3249-3317 19 7.0 H C2 30 3254-3309 19 3.0 R C2 31 3309-3357 19 3.0 R C2 32 3189-3259 18 3.0 R Cl, C2 Notes: 1 R = room temperature; H = heated sample (160 C) 2 Cl = CO (1,0); C2 = CO (2,0) 180 2.18% 12Cl3CH ; and o.01% 13C 13CH2 (Barnes, 1972). Gas pressures have been measured with MKS Instruments (Burlington, MA) Baratron gauges. For pressures less than 1.0 torr a 0-1 torr head was used while pressures greater than 1.0 torr were measured with a 0- 100 torr head. A model 170M-26A meter unit, a 170M-7A electronics unit, and 170M-34B multiple head selector head were used. The calibration of the pressure readings was viewed as very impor tant. Both gauge heads were sent to the factory for calibration prior to beginning the measurements. Approximately midway through the inten sity measurements both heads were again sent to the factory to have the calibration checked. The readings of the two pressure heads were also intercompared throughout data acquisition. At pressures greater than 0.1 torr, the readings were always found to agree to within 0.5%. Pres sure readings were compared on several occasions with those obtained with Baratron gauges owned by the National Oceanic and Atmospheric Administra tion/National Environmental Satellite Service. Since the NOAA/NESS gauge was calibrated at the National Bureau of Standards, an independent check of the pressure readings was obtained. Within the range of pres sures used for the intensity measurements, the gauges agreed to within 1%. In Figure 38 pressure readings obtained with the Ohio State and NOAA/NESS Baratron gauges are intercompared at two epochs. The procedure used to fill the sample cells is as follows. Sample cells were pumped for two or more hours with a diffusion pump equipped with a liquid-nitrogen-filled cold trap. When the system pressure reached approximately 2 x 10 5 torr (as measured with a Varian model 524-2 cold cathode gauge and model 860 meter), the zero points on the DIFF (Pt.-Poso/Pt) x 100 .f o 2.of- -2 .L O l.0L Figure 38. Comparison of pressure gauge readings obtained with the the with obtained readings gauge pressure of Comparison 38. Figure - A O ° O OCTOBER 77 OCTOBER O ° O ° O ° A AU A A U A A A 0 20 0 40 0 60 0 80 0 1000 900 800 700 600 500 400 300 200 100 were taken with the Ohio State 0-100 torr head and the the and head torr 0-100 State Ohio the with taken were between the readings of the two gauges has been plotted plotted been has gauges two the of readings the between NOAA/NESS 0-1 torr head. The percentage difference difference percentage The head. torr 0-1 NOAA/NESS here data The Baratrons. State Ohio the and NOAA/NESS vs. the pressure in pm Hg. pm in pressure the vs. s O © G © O O ” s 5 P(/zmHg) RSUE AG COMPARISON GAUGE PRESSURE A 78 MAY A NOAA- OSU NOAA- A A 181 182 Baratron gauges were checked and adjusted, if necessary. The pumping system was then valved off, and the acetylene gas was allowed to flow slowly into the system until the desired pressure was reached. The system pressure was monitored for several minutes to assure that equili brium had been reached. At this point the pressure and cell temperature were recorded and the cell valve closed. The sample cell was immediately taken to the spectrometer to record the data. Cell temperatures were continuously monitored with precision ther mistor probes and a model 46 meter obtained from the Yellow Springs Instrument Company (Yellow Springs, Ohio). The YSI Series 400 probes have a quoted accuracy of ± 0.1 C while the meter accuracy was stated to be ± 0.15 C. Nine glass cells having path lengths ranging from 1.988 to 96.95 cm were used for the intensity runs. For the longest cell the path length was measured with a Pratt and Whitney 80-inch traveling microscope. Measurements were taken at 90 degree intervals along with circumference of the windows, averaged, and the CaF x<7indow thicknesses subtracted to obtain the gas path length. The readings at the four points agreed to within 0.01 cm. The path lengths of the shorter cells were measured to an accuracy of 0.003 cm with precision calipers. System linearity was tested by placing a sapphire window in the beam, and the fractional decrease in signal amplitude was measured as a func tion of source intensity. The fractional decrease in the recorded out put was found to be independent of the beam intensity. By comparing the signal level at the center of a saturated line with that obtained by blocking the beam at the exit slit and placing the prism between orders, 183 the amount of scattered light was found to be less than 0.3%. Scattered light corrections were not applied to the intensity data. Long term source-electronics stability was tested by running the chart recorder at a fixed grating position for about one hour. No drift was noted during that period. The analog records were digitized with a Bendix Datagrid Digitizer. Approximately 2000 data points were recorded per line by moving the cur sor along the spectral line profile. A linear least squares fit to the data points on both sides of the line was used to determine the zero absorption level. About 0.10 cm-1 of spectrum was included in the least squares fit. Successive measurements of the equivalent width of the same record agreed to better than 0.5%. Accuracy was limited primarily by the ability to follow the spectral profile manually with the cursor. A digitized scan is shown in Figure 39. A total of 709 intensity scans were made, all at room temperature. For each line four or more scans at between two and eight pressures were made. For each scan the zero level was measured prior to and after the run by moving the prism between orders. Calibration lines on both sides of the line being measured for intensity were included to calculate the dispersion. The use of goniometer fringes increased the accuracy of the dispersion measurements. The intensities reported are the mean values of all runs at 300 K. The pressures and path lengths were adjusted for each line so that the equivalent width was between 0.003 to 0.010 cm-1. For the experimen tal conditions this range of equivalent width is very close to the linear 0.0 0.0 0.2 0.3 0.4 CT m 0.5 x X 0.6 O.S.U. DIGITIZED OUTPUT: f LIMITS OF INTEGRATION 0.9 2320. 5 I 2320. 61 2 3 2 0 .GG2320.56 WAVENUMBER Figure 39. Profile of an absorption line obtained with the Ohio State University 10-m focal length Czemy-Tumer vac uum grating spectrometer. The data were digitized with a Bendix Datagrid'digitizer. Arrows mark the limits of integration used to determine the equivalent width. 185 part of the curve of growth. Pressures were kept as low as possible to minimize the effects of collisional broadening. 12 Section 6.5 - Partition Function for C2H2. In order to calculate the dipole moment matrix element from the observed line intensity, it is necessary to know the total internal partition function. This quantity allows the relative population of an individual rotational level to be calculated at temperature T under the conditions of thermal equilibrium. The partition function can be used to relate the observed line strength at one temperature to the line strength at any other temperature. All thermodynamic quantities can be expressed in terms of it (Herzberg, 1945). In the present study the total internal partition function has been 12 evaluated from a tabulation of the molecular constants of all ^ 2^2 -1 vibrational levels within 2200 cm of the ground state. The rotational and vibrational constants were assembled from several sources (Scott and Rao, 1965; Palmer, Mickelson, and Rao, 1972; Pliva, 1972a; Pliva, 1972b; Baldacci, et a l ., 1977) and are shown in Table 23 along with the appro priate nuclear spin statistical weight factors for the rotational levels of each vibrational band. The rotational constants for many of the vibrational levels listed in Table 23 must be regarded only as effective constants because of the effects of 5,-type resonance. The calculations were performed with Equation 6.3-6. For each vibrational state, rotational levels up to J = 100 were included. Values of Q(T) for the temperature range 200 K to 350 IC are tabulated in Table 24 at 10 K intervals along with the fractional population of Table 23 1 8 6 R0-VIBRATIONAL CONSTANTS (cm”1) AND STATISTICAL WEIGHTS FOR LOWER LEVELS OF 12C2H2 NSTATE SYMMETRYV 3 a J°EVEN J=0DD 0 D M 0 6 HMO 10 1 0000°0° l+ 0.000 1.17660810.000014 1.61010.007 0 1 3 g 01 O 1.17533910.000024 1.64010.015 1 3 o o o 2 O n 612.870i0.002 l g 1.18055810.000025 1.65410.017 3 1 1.17641210.000018 1.61010.009 3 1 3 0000°l1f n 730.3314±0.0016 l u 1.18111210.000019 1.65510.010 1 3 4 0002°0° z+ 1230.39i0.05 1.179510.0007 -8.514.1 0 1 3 g 1.177210.0003 7.410.8 1 3 2 0 5 0002 0 A 1233.5210.05 2 g 1.178410.0003 3 1 6 000(11) ° Z* 1328.073510.0017 1.1805010.00004 3.6210.03 3.0810.14 0 u 3 1 7 000(11)_° l" 1340.54710.003 1.1801010.00003 1.6910.02 0 u 1 3 1 e 1.1798510.00006 -0.20l0.Q5 -2.7+0.3 3 1 a 000(11)" A 1347.52410.005 u 2 1. 1798410.00005 1.6210.02 1 3 + 9 0000°2° I 1449.12110.003 1.1811010.00003 4.1010.02 4.9H0.05 0 1 3 g 1.1807410.00003 -1.07+0.04 -6.710.2 0 2 e 1 3 10 0000 2 A 1463.01310.003 2 g 1.18077+0.00003 1.6710.01 3 1 1 0 e 1.178 1 3 11 0003 0 n 1855.7210.05 g 1 1. 163 3 1 3 0 e 1.184 I 3 12 0003 0 1861.9310.05 g 3 1.184 3 1 1.1779510.00005 2.8710.12 1.810.9 3 1 13 000(21)1I1^ n 1941.17910.003 u 1 1.1856910.00005 3.1010.10 0.410.6 1 i 1.1826510.00008 3.4610.29 8.612.8 3 1 14 000 (2 l)11 6 n 1960.B7410.004 1 f u 1.1799010.00007 3.2210.25 5.912.3 1 3 1.186 3 1 15 000(21)3 ® 4* 1962.1 u 3 1.186 1 3 16 0100°0° Z+ 1974.31710.003 1. 1704010.00003 1.6510.03 g 0 1 3 1.1778310.00004 2.3510.04 0.910.1 1 3 17 000(12) 1ixe n 2049.05910.009 f g 1.1874910.00003 2.6610.04 1.010.2 3 1 1.183 1 3 18 000(12)1I e n 2066.998 f g 1.180 3 1 1.188 1 3 19 000(12)3 * 4 2085.5 g 3 1.188 3 1 0 1 G 1.1785810.00003 2.5110.09 1.510.6 3 1 20 0000 3 JI 2170.34310.002 u 1 1.1881510.00003 2.7010.08 -2.310.5 1 3 0 3 e 1.189 3 1 21 0000 3J 2198.1 u 3 1.189 1 3 TABLE 24 12c 2H2 ?ARTITI0N FUNCTION Vibrational Band 1 2 3 4 5 6 7 8 9 T Q(T) 10 11 12 13 14 15 16 17-21 200.0 245.4252 0.9657 0.0235 0.0100 0.0001 0.0003 0.0001 0.0001 0.0001 0.0000 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 210.0 259.9259 0.9573 0.0287 0.0128 0.0002 0.0004 0.0001 0.0001 0.0002 0.0000 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 220.0 274.9698 0.9479 0.0344 0.0159 0.0003 0.0006 0.0002 0.0001 0.0003 0.0001 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 230.0 290.6077 0.9376 0.0405 0.0194 0.0004 0.0008 0.0002 0.0002 0.0004 0.0001 0.0002 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 240.0 306.8898 0.9264 0.0469 0.0232 0.0006 0.0011 0.0003 0.0003 0.0006 0.0002 0.0003 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 250.0 323.8649 0.9143 0.0537 0.0273 0.0008 0.0015 0.0004 0.0004 0.0008 0.0002 0.0004 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 260.0 341.5815 0.9015 0.0606 0.0316 0.0010 0.0020 0.0006 0.0005 0.0010 0.0003 0.0005 0.0001 0.0000 0.0000 0.0000 0.0000 0.0000 0.0001 270.0 360.0871 0.8880 0.0677 0.0362 0.0013 0.0025 0.0007 0.0007 0.0013 0.0004 0.0007 0.0001 0.0000 0.0001 0.0001 0.0001 0.0000 0.0001 280.0 379.4288 0.8739 0.0749 0.0409 0.0016 0.0031 0.0009 0.0009 0.0017 0.0005 0.0009 0.0001 0.0001 0.0001 0.0001 0.0001 0.0000 0.0002 290.0 399.6529 0.8593 0.0820 0.0458 0.0019 0.0038 0.0012 0.0011 0.0021 0.0006 0.0012 0.0002 0.0002 0.0001 0.0001 0.0001 0.0000 0.0002 300.0 420.8057 0.8442 0.0892 0.0507 0.0023 0.0045 0.0014 0.0014 0.0026 0.0008 0.0015 0.0002 0.0002 0.0002 0.0001 0.0001 0.0001 0.0003 310.0 442.9326 0.8287 0.0963 0.0558 0.0027 0.0054 0.0017 0.0016 0.0032 0.0010 0.0019 0.0003 0.0003 0.0002 0.0002 0.0002 0.0001 0.0005 320.0 466.0791 0.8129 0.1032 0.0608 0.0032 0.0063 0.0021 0.0020 0.0038 0.0012 0.0023 0.0004 0.0004 0.0003 0.0002 0.0002 0.0001 0.0006 330.0 490.2901 0.7969 0.1100 0.0659 0.0037 0.0074 0.0024 0.0023 0.0045 0.0014 0^0027 0.0005 0.0005 0.0003 0.0003 0.0003 0.0001 0.0008 340.0 515.6101 0.7807 0.1166 0.0709 0.0043 0.0084 0.0028 0.0027 0.0052 0.0017 187 0.0032 0.0006 0.0006 0.0004 0.0004 0.0004 0.0002 0.0010 350.0 542.0835 0.7644 0.1229 0.0758 0.0048 0.0096 0.0032 0.0031 0.0060 0.0020 0.0037 0.0007 0.0007 0.0005 0.0005 0.0005 0.0002 0.0013 188 each of the 16 lowest vibrational levels. Polynomial coefficients in T accurate to the seven significant figures over the same temperature range are presented in Table 25 (powers of 10 are in parentheses). Section 6.6 - Analysis of the Spectrum at 3 pm This section presents an analysis of the bands observed in the absorption spectrum of acetylene in the 3 pm region. The primary objec tive of this project was to identify "hot" bands arising from the 0002° ,20°, 000(11)°’2 and 0000°2°’2 levels which have sufficient intensity to be detectable in stellar spectra. In addition, corrections in the rotational assignments were made in a few instances for bands previously reported in the literature. 12 The strongest absorption in ^ i s from the bands and + (v. + vr)° , , which are in strong Fermi resonance and of comparable I a 3 /A- intensity. A total of five IT-H bands from the 0001°0° and OOOO0!1 levels were identifed and are of moderate intensity in the room-temperature spectrum. In the present work it has been possible to identify 14 12 additional "hot” bands of ar^s-‘-n§ from higher vibrational levels and 5 bands of 12C°3CH^ from the ground state, 0001°0°, and 0000° l1 levels. The 27 bands identified provide data on 41 different vibrational states. Figure 40 shows a short section of the 3 pm spectrum recorded at room temperature and a pressure of 3.05 mm Hg, with rotational assign ments given for the stronger lines. The strongest absorption features are from the v„ and v0 + (v, + v')?,. bands. Lines from several II—II 3 Z b 2,“}- bands are indicated and exhibit resolvable rotational L-type doubling. 189 Table 25 12 C2^2 Partit:*-on Function Polynomial Coefficients a = -1.8923462300 (+01) o 1.6578377395 (+00) al = -4.1672521882 (-03) a2 = 1.4106025215 (-05) a3 = 11 D3 D3 -9.4194026521 (-09) ■F* i ■F* 5.2942462726 (-12) a5 = Q(T) = aQ + a1 T + a2 T2 + a3 T3 + a^ T4 + a5 T5 ABSORPTION iue4. ctln setu ewe 39. n 39. cm 3291.3 and 3290.1 between spectrum Acetylene 40. Figure y -ye obig Te nest ai i 31 for 3:1 is ratio intensity The doubling. £-type by eodda ro eprtr ihapesr o 3.05 of pressure a with temperature room at recorded ie of lines stronger The cm. 96.95 of length path a and torr ttsis a e en o oainl ees split levels spin rotational for nuclear of seen be effect can The statistics identified. are features 3290.5 3291.0 190 191 Components with the same J value occur with an intensity ratio of 1:3 because of the effects of nuclear spin statistics. Although only a natural sample was used, rotational structure of the V3 band of is of moderate intensity in the spectrum. Short sections of spectra recorded at room temperature and at 160 C are compared in Figure 41. In addition to the strong resonance betx^een v and v (v + v )2., b 3 2 i+ 5 Fermi resonance is expected x^ithin a number of higer-energy diads and triads. Calculated unperturbed energies are presented in Table 26 for a number of resonating polyads expected as upper levels of transitions observable at 3 ym. From the data it can be seen that a very close resonance is expected between the levels in columns 1 and 3. Since the uncertainty in the unperturbed energies is likely to be large, it appears to be impossible to distinguish the ordering of the levels in columns 1 and 3 at the present time. In this analysis the convention has been adopted that the levels in column 1 (with v^ = 1) are assumed to have higher energies than those in column 3. In reality, a strong mixture of wave functions of the interacting levels will occur, owing to the extreme nearness of the unperturbed levels. Molecular constants reported in Table 27 and 28 were determined with the polynomial in "m" program of Palmer (1972). The errors quoted are the standard deviations in units of the last significant decimal place. Because many of the levels analyzed in this study are perturbed by A-type resonance and/or Fermi resonance, the rotational constants obtained for many bands are only effective values. For these transitions extrapolation to J values outside the range of the measured values will produce large errors. For two highly perturbed bands, systematic rather cfff' 3207.00 3208 00 320900 3270.00 Figure 41. Comparison between acetylene spectra recorded at 12.0 mm Hg pressure and 160 C (upper) and at 3.5 mm Hg pressure and room temperature (lower) in the region 3207 to 3211 cm-1. 193 Table 26 Calculated Unperturbed Energies for Several Diads 12 -1 and Triads of ^2^2 *n ^erm^ Resonance (cm ) Level G Level G Level G o o o 0010°0° 3288.AO 010(11)° 3288.34 001110° 3890.27 010(21)^ 3888.97 010(21)* 3908.66 00100!1 4007.84 010(12)^ 4007.76 010(12)* 4025.69 0012°0° 4496.79 010(31)° 4495.91 “T 001220° 4499.92 010(31)^ 4496.84 010(31)2 4524.99 001(11)° 4594.58 010(22)°IT 4594.24 010(22)°x 4630.26 001(11)° 4607.06 010(22)° 4607.41 OOl(ll)2 4614.03 010(22)^ 4612.37 010(22)2 4638.79 0010°2° 4715.74 010(13)° 4714.93 0010°22 4729.63 010(13)2X 4730.32 010(13)2 4752.63 T.iMi* ii Mtilucnl;ir Const.ints ( ) of lin Per I veil from the l'aruls .it 3 Urn .IMAX St. Dev. vo-i;1 v r r (n,-B")xin1 (n,-n")xio6 (ii'-iOxio10 (i.'-L")xio1A UPPER V R xlO3 0 010(11) o: '*> onoo°u° O' +) 11:81 .0020 i 6 -4.099 3- 4 0.853 3 7 1 . 92 ± 5 2.25 ± 9 52 51 2 .4 u g J269. 5399 ± 10 -4 . 676 3 5 0.515 -l 6 0.60 3 2 47 43 3.5 010(21)1 (IT) (II ) 0001!00 (II ) f-f 32 09. 3436 i 'J -3.127 ± 4 0.797 3 6 0.88 ±2 48 44 3.2 e-n 3272, 1050 ± 10 -5,164 t 6 0.457 - 8 0.56 15 46 42 3.9 010C12)1 (IT) (n ) 0000°1] (N ) f-f 3272. 1071 ± 10 -2.421 i 6 1.075 ± 7 1.33 ± 3 46 43 3.8 010(31)° (IT) O'. +) 0002°0° (E +) 3258.4594 i 35 -4.480 1 40 -2.138 ± 132 -21.62 ± 1.59 -64.51 i 6.46 39 33 8.0 e-e 3257. 2984 i 30 -3.071 2 41 3.977 ± 155 36.01 ± 2.26 143.71 ± 11.71 37 29 7.1 010(31) (TT) (A ) 000220° (A ) f-f 3257. 3180 i 18 -3.767 ± 8 0.489 ± 7 36 34 5.9 0 010(22) a B ) 000(h) - o; u ) 3259.2282 ± 15 -3.847 ± 9 0.493 ± 10 33 29 5.6 e-e 3260. 4 7 30 i 30 -3.639 3 36 -1.422 ± 102 -15.76 ± 78 31 28 7.2 010(22) (II) (A ) 000(11) (A(1) f-f 3260. 4755 1 19 -3. 625 9 0.548 ± 10 37 28 6.6 (110(1 3)^ (II) (r. +) noonn2° a 3 126].6321 3 22 -3.676 i 23 2.147 t 47 22 20 4.2 e-e 3262. 8331 ± 35 -3.738 ± 48 -1.916 ± 176 -16.36 ± 1.83 27 23 6.2 (A ) 0000°22 (A ) f-f 3262. 8332 ± 22 -3.682 ± 12 0.583 t 14 29 30 5.7 0 0 1 0 °0 ° (1 +) oooo°o0 O' +) 3294.8406 i 9 -4.213 ± 6 1.658 ± 10 4.65 ± 6 5.02 ± 12 53 51 3.5 e-e 3205 ,4715 ± 11 -4.662 1 6 0.900 ± 7 1.45 ± 2 47 46 3.9 oouV in ) onoi1 o° (ii ) f-f 3285 .4625 i 12 -4.082 ± 6 1.038 3 8 1.19 ± 3 46 46 4.4 194 Table 27 (cent inued) l.EVKLS JMAX St.Dev. ,v ,7 v0-n' v 2+n' ixlO1 (I)’-l)"]KlOr’ (H’-[lM)xl0lf} (L'-L*W 4 UPPER LOWER r R xlO3 e-e 3286.3846 t 10 -5.676 * 5 0.176 " 6 0.59 1 2 4 6 48 3.8 ooio0!1 ooooV (II ) u < V f-f 3286.3868 ± 11 -3.588 i 9 -0.316 ‘ 21 -5,06 " 17 -7.59 ± 42 46 44 3.5 nai 2non (T + ) n 002n 0° 32 77. 63RO J- 36 -5.184 i 63 1 .067 272 20.41 i 1.97 111.35 i 18.53 35 31 10.4 1) e-e 3275.704 7 ± 51 -3.311 ^ 69 1.046 ± 259 -3.01 ± 3.50 -32.13 37 32 12.6 2 0 ± 15.54 0012 0 (A ) 000220° 11 001(11)" a + ) 000(11)° 3281.2629 ± 21 -4.84 5 ± 32 2.081 ± 110 15.32 ± 1.00 28 26 5.7 g O a 3277.3741 ± 14 -4.485 ± 10 0.701 ± 12 30 28 4.8 001(11)° V 000(11)° u 3277.2561 ± 22 -4.901 ± 24 -1.560 1 65 -6.12 ± 47 30 30 5.0 001(11)2 000(11 )2 (A ) 11 0010°2° e-e 3239.7258 ± 14 -7. 728 ± 6 0.020 ± r» 38 28 4.8 inoiV (n ) nooo0!1 tn ) p. f-f 3239. 7295 ± 1 3 -7.115 ± 6 0.009 ± 5 38 3L 4.5 0110°0 ° (E +> 0100°0° (£ +) 3285.7058 ± 36 -5. 599 ± 3(1 -0. 350 + 71 25 19 6.8 195 Table 28 -1 12 13 Molecular Constants (cm ) of C CII^ Derived from the Rands at 3 pm LEVELS JMAX St.Dev. V o-B'£.'2+B''L"2 (B'-B")xl03 (D'-D")xl06 (U'-H")xl010 UPPER LOWER P R xlO3 0010°0° (Z+) 0000°0° (Z+ ) 3284.1904 ± 7 -5.499 ± 5 -0.015 ± 7 0.42 ± 2 42 44 3.1 e-e 3274.7079 ± 16 -5.554 ± 9 -0.109 i 10 31 30 3.8 0011l0° (n) oooi1o° (II) f-f 3274.7045 ± 19 -5.283 ± 15 -0.088 ± 27 30 19 4.2 e-e 3276.8239 ± 34 -5.550 ± 23 -0.121 ± 29 27 29 8.4 00100!1 (n> 00000 !1 (n> f-f 3276.8367 ± 33 -5.611 ± 22 -0.174 i 25 29 28 5.7 (01011)° (£+ ) 0000°0° (I+) 3250.4842 ± 20 -2.715 ± 16 1.361 ± 31 30 18 3.9 1000°0° a +) 0000°0° a+) 3361.5759 ± 20 -6.424 ± 30 -0.024 ± 104 16 20 5.8 196 197 than random differences exist between the observed and calculated values even with an eighth order polynomial. Observed and calculated line positions (cm "*) are presented in Tables 29 to 67. Vibrational term - 1 12 values (cm ) of C2H 2 obtained in this study are presented in Table 68. Transitions from the Ground State Bands of 12C2H2 : Both the v 2 and v 2 +(v^ + ^ 5) ^ bands have been measured up to approximately J = 50 in the present work. A satisfactory fit to the observational data required an eighth-order polynomial in both cases. The need for additional terms can be attributed to the effects of Fermi resonance and 5,-type resonance in the upper states. The lower-level combination differences were found to be in agreement with values com puted with the ground state constants of Palmer, Mickelson, and Rao (1972). Bands of 12C 13CH2 : Although only a natural sample was used in this study, it was possible to measure lines of the v ^ band up to J = 44 in the R branch and J = 42 in the P branch. A sixth-order polynomial was required to fit the data accurately. The values obtained are in agreement with the constants of Lafferty and Thibault (1964) and Ghersetti, et_ al. (1975). The ground-state constants of this molecule were calculated by fitting the AF''(J) values obtained in this study and in several other studies (Baldacci, Ghersetti, and Rao, 1973; Ghersetti, et al., 1975; Baldacci, Ghersetti, and Rao, 1977) to the standard expressions. The rotational constants B = 1.148415(18) cm“ ^ and D = 1.550(11) x 10 ^ cm 1 obtained are in good agreement with the values of Lafferty and Thibault -1 198 table 2 9 0 observed aimd calculated waveimumber^ svaloCm j 0 0 o„o 12 1 3„1t SjF T he 0 0 1 0 0 - 0 0 0 0 0 bAND OF c ch2 J M u f 0 b S R SJ{CALC 0-C PS JDOBS PS J } GA LC a-c AlCO X 1 0 0 0 3286 o4?4 2 2 8 6 o4 7 6 “ 0 a 2 A 3 2 b 8 o7 3b 3 2 8 8 o 0 3 2 0 03 328 Ao 892 3281o893 ” 0 0 2 2 32 V1 0 017 2 2 9 1 o 0 15 0 0 1 327 Vo 3b8 3 2 7 9 0585 0o3 u 32 93 0 3 69 2293o26b Ool 32 77 0 265 3277 0 266 “Ool -n 3293 o 310 3295 0 310 » 0 o0 3274o 935 32 ho 937 “ 0 0 2 y 32 9 7 0 043 3297 0 7 4 1 0 a 2 327 2 0 3V3 3 2 7 2 0 39 6 “ Ool 0 3299 0 93b 3 2 9 9 0 9 b 0 = 0 o4 3 2 7 0 o 2 4 i/ 327 0 0 244 0 0 3 1 3302 0 Ibb 3302 0 Ibb “ 0 o 2 3 2o7o 884 3267 o882 0 0 2 b 3304 0 3b4 3 30 4 0 26 3 “0 G 1 326 3o 507 3 2 6 5 0 309 -0 0 3 9 •2306c. 340 3306 0 550 “ 0 0 4 3263o A26 5 2 6 3 0 126 OoO 1 0 33U8 o726 330b o?2 4 0 o2 3 260o728 3260o 731 “ 0 0 3 1 1 3 310 0 fa 8 7 323 bo 329 3 2 5 8 o 327 0 0 2 A2 3313 0 040 33i5o 038 0 a 2 3 2 5 3 0 9 A1 32 3 5 0 V A A “ OoO 13 2 3 a 3 oil 77 3 3l 3 0 a ? 8 “Oal 325 3o 486 3253c-485 Ool A4 2317 0 3 lU 3 317 0 2 0 6 0o3 325lc052 3 2 3 1 0 049 0 0 3 13 23 A V O 424 3 3i9 o423 Ool 3 240o 595 3248 0 602 ” 0 o ? it. 3 3 2 1 0 330 3 3 2 1 o 5 2 V Ool 3240 0 146 3246 0 143 OoO 1? 3 3 2 3 0 624 3 3 2 3 o 623 Ool 3243ofcb3 3 2 4 3 o 678 0 0 4 ie 332 3 0 707 3 3 2 3 o 703 Ool 3 2 4 l o 204 3 2 41 0 2 01 0 a 4 i.9 3 3 2 7 0 7 7 4 332? 0 77 o “0 0 3 3 2 3 b o 7 l 6 3 2 3 8 o 7 13 0 0 3 2 0 3 3 2 9 o 636 5329 0 8 2 6 OaO 3 2 3 6 o 2 16 3236 0 216 U O’ 0 2 i 333 A 0 bb 3 3 3 3 1 0 8 8 4 “Ool 3 266 o 7 0 i 3233o 70b “Oo? 22 3333 0920 3 3 3 3 o92 I “ Ool 3 2 3 l o 188 3231 a 190 “0 0 2 2.3 3335 0 943 3335 cV4b “ 0 c 3 322 bo 662 3228o663 “Ool 24 3337 0 962 5337c. 960 Oo 2 322oo127 32 2 6 0 126 Ool 2 3 333Vo958 3 339 o 963 “0o5 3223o 5 7 o 32 2 3 0 37 9 “ 0 c 3 Zt> 3 3 4 1 0 934 3 2 2 1 0 026 32 21o023 0 0 3 0 z l 3343 0 937 3 34 3 0 93 3 0 f4 3218 04b7 z t 5 34 3 o904 3 2 1 3 o 879 3215 0 8 83 “ 0 0 3 2V 3347 0 8 6 O 3347 0 862 “ 0 c 2 3213o 297 3213o299 -Ool 30 334 9 0 B03 3 3 4 9 0 8 1 0 “0 0 4 3 2 1 Oo 707 3 2 1 Oo 7 06 0 o2 31 335 A 0 75 A 335 A a J4b 0 o3 320bo108 3208ol04 0 0 4 3 2 5363 o672 320 3 0 496 320 5 a 4 94 0 0 2 33 3 3 3 3 0 369 2233 o368 Oo A 3202 0 bOb 3 2 0 2 o 8?3 “ 0 0 6 J>*4 3337 0 49b 3337 0 4 9 4 0 0 ^ 3200o 2 5 0 3200 0 248 0 0 2 33 3 3 5 9 0 3v3 3 3 5 9 o38V 0 0 4 3iV7o 615 3 1 9 7 o 6I3 Oo 3 36 3 3 o 1 0 2 73 31V4o9fa9 3 A94 0 V70 -Ool 37 3 3 b 3 o 150 3363 0 131 “•0 0 1 3 A9 2 0 317 3192 0 320 “ 0 0 3 3b 2363 c.020 5365 oOlb 0 o2 318 9o 663 3189 0 6 6 2 Ool 39 3 3 0 6 0 a7 i 3 3 o 6 a b 7 3 “ 0 0 5 3 A b 0 o 9 9 3 3lb6o997 “0 o2 40 3368 o 724 318 4 o 329 3 184 o 3 2 6 0 0 3 **■1 3370o 5b3 3 3 7 0 o 5 b 5 "0 0 2 3 1 fc1 0 649 4 2 3372 0 393 337 2 0 398 “0o5 3 A7 8 0 963 3178 0 963 -Ool 43 337 4 0 2 1V 5 3 7 4 o 223 “0 03 ^4 337 6 0 048 5376 o040 O 0 8 ~1 199 TAdLt 3 Oo DtJScKVtD AND CALCULAlfcD WAtftNUMb£RS 4Va C.CM > 1 0 0 1 12 UF InE 1001 o - oooc>1 bANL) UF e e L 2h2 vjl Ki3 SUbi. k 13 a c a l c G-t P 5 J{Obb P5 J SCA l C 0-C X I 00 xlOO 1 1 1 1 f** 0 0 1 3244 o 3 93 3 244 o 3b 6 2 3 240 0692 323bo002 323 5 0 004 — 0 c 2 3 232o620 0 3 3248 o90 3 O 1 0 4 32 b1 o 26 b 32bi o259 O 323 Oo 217 32 30 02 21 3 2 2 ? o 806 O b 32b3obA 8 0 1 0 3 >«e,ca«s«» “ «■»«■»«!» «b»«s> 0 «=»«» ■™«» J K 4J50bi> K id 5CAL0 0-C Pi JiDBS Pi J 5 CALC 0-C X 100 xioo I 3244 o4 11 3244 o^il “ 0 o0 £ 324-6 o730 323 5o 002 32 34 o 9 91 1 a 1 3 3249o030 3 2 4 9 o 03 5 —0 a 6 323 2 0 60 A 32 32 0 601 OoG 4 3 2 5 1 o 32 6 3230o 2 92 3 2 3 0 a 196 “0o4 5 3253 o602 3227 0 77 8 6 02 35 o b 66 3255 o b6 3 0 o 3 2225 0345 I 323 6 a 108 3258 oil 1 —0 o 3 322 2 0 895 322 2 0 898 — 0 a 3 c 3260o339 32b0o343 “0 a 4 3220o438 9 0202 o 537 3262 o56 1 “0 o 4 3 2 1 7 0 966 32 a? 0 963 0 o3 i.0 6264o?62 3 OF iMt 0 0 1 1 1 0 ° - 0 0 0 1 1 0 ° bANU OF 12 C H e e 2 2 F< 13 JOBS f\ 4 *ji J LA l C 0-C Pfl J}UBS P«J JCAl G U-C XiOO X100 1 3 2 9 0 o 14 b z 3292 o46b 32 BO 0 7b 1 3 3 2 9 4 o 776 3 2 9 4 o 781 -0ob 327 b 0 3b? 3278 ©391 “0 o4 UF The 001110°_ - 000110° _ BAND OF 12 L M f f 2 2 J R I J $ Utob K tJ 5C alC U“ C PIJ iUbb F1J }CAlC U“L X100 A IOC A 3 2 VO o16b 32 VO0AbO Oob z 32V2 c.497 3280c732 3 3^VMob2V 3294ob2b Oob 327 bo56b 5278 o355 OoO 4 32 97 o 14 9 3297 0 144 Oob 327boV78 3 2 1 bo 969 0 o 8 b 5 2 9 9 o456 3 2 9 9 o4b5 0 0 2 32? iobbA 3273o b ?6 Oob 6 3 3 0 A o 754 330io7b 7 •“0 0 2 32? Ao 17b 5 2 7 1 0 17m 0 0*9 ? 3 3 0 4 o046 3304o049 “00 A 3268o?bb 3268 o7b4 0o 39 3370 0 b 10 3370 a 518 “ 0 0 b 5Abbob34 mO 3372 0 303 3372 0 308 “ 0 a 4 31b3o003 *0 A 3b 7 m0 07 1 3374oQ7? “ 0 c 6 3i@ 0o155 m2 337b 0 b2b 33 #5 0 8 2 b “0 0 A 3 17 7 Q 285 317 7 0 2 9 0 •“Oob m3 3377 0bb1 33/7o554 “ 0 o3 3A?4o407 MM 337 V 0 2 bO 5379 0262 “ 0 0 2 3 a7 lo 507 3 1 7 1 0 507 Ooi Mb 3380o957 3380 09 4 V O08 3 lb bo 59 m 3Abb obb9 Oob Mb 338 2 0 b2 0 3 362 0 6 16 0 oM 3 Ab bo 1 53 3165o655 “Oo 1 - 1 204 1 ABLE 35 o. UbdE ktf E D AhU CALCULAT El) i-M\/E(MUr-1BEKi. J VAC oCM } UF THt 00100!1 - OOOO0!1 bAlMD OF L H e e 2 2 OKS.e.OCWBOCTOeBWO CioaCS <=«=-==.<=<=■ «=* c=»c=.«yM= .:-jii T-n.r:-.-«=»•.a«aoea*»'CS<=3a «=.*=. J R d j D U B S Rto DCAl C 0 “L Pt JD Ubi Pt JSCALC 0 “L XiOO ;aoo 5so '—* c>c« cJ> e^ca c 1 32Vio060 329A o066 0 ob 2 32V3o3?5 3293 037 to - 0 o0 3 2b 1 o 662 3 2 8 1 0 6 6 ? “ 0 0 6 to tolVbototoO 3ZV5o683 “ 0 o 2 32?Vo28l 32?9o292 ” i 0 i H toivioVil 329? o9 ? 8 “ Ool 327 6 o 901 tollb aV 0 b “ 0 ob 6 3300 o268 3300 o 262 •==0 o <6 toll Ho 600 327bo 607 “ 0 0 6 to totoOloto ton 33Qzo53b OoO toll lo 100 3z 7 2 0 09? 0 0 3 1 totoOHolV? 3 30djo tfV6 Oc.3 3269 o 677 to 3307 o 0b3 330 ? oO8) 3 0 o0 326 ? a m i 326? 0 2b5 0 0 2 V totoOto oi l n 33 0 V o 2 ? V “0 o 6 326bo 802 aO totoilo30V 3 3 A1 o503 0 o 6 326 2 o3b 6 32 6 2 0 3b? “ 0 0 2 il 3313 o 7 17 3 3 A 3> o ? 1to 0 o 3 3Z6Vo8dI 3 2 6 90 88 2 “ OoO 12 btoltooVito 3 3 A 5 o v 1 b Ool 3 2 6 7 o bQ b 3257ob0b Oo 3 13 33ibolO 3 3318o100 0 06 325 bo 918 326 b o 9 16 0 0 2 A b totoiOoiib 3 3 i 0 oi l n 0 o 3 3252obib 3262obl6 0 0 2 lb totolloHJto 3 tollo4 3 to “0 a I 32b Vo 905 32bVo 90b Ool 16 totokno6B3 to'lnl o381 3 2 b 7 0381 0 o0 l i 3326o7IV 332too ? 1 1 0 c 2 3 InnabtoO 32bb 0 8 b6 0o5 11 332 b o 8 b 2 3 3zdo d3 b Oo 3 32b 2o 302 32b2 0 29V 0 0 3 19 333 0 o Vb? 3 330 o 9b6 Ool 323Vo 738 3239 0 7b0 “ 0 0 2 20 3333o 038 3333o039 “ 0 o 1 3 2 3 ?o17b 3 2 3 1 0 1 1 0 Oob Zl 3 3 33oiIV 3 2 3 4o3V0 323 b 0 58? 0 0 3 333 ? o 16b to 33 1 o 18 6 “ Ool 3 23 lo 991 3231 0 9 9 2 “0 0 2 333V o 2d>I 333Vo237 0 o 6 322 Vo 379 322 9o 3b 6 “0 o? 33d» 1 o 11 n 3 3b3LoZ?b OoO 322bo 770 3226 0 766 0 0 3 lto totontoolVt 3 3 da-2> o 2 9 ? 0 o 2 3llHo lb 2 32 2 b 0 136 Ool lto totonb & zV V 3 3b6o3 05 “ 0 a 6 3 2 2 lobVl 3 z 2 1 obVl 0 o 0 l ? totoHlo300 33b7oZ98 0 o3 3218 u 8 2 V 3 2 18 0 83b “0 g 5 26 33d) V o 270 totonVolfib “0 06 3 z l 6 o JL6 b 3216 0 16 5 “Ool IV 333 A o zd) A to 3 to 1 o 2 3 V Ool 3 2 l3 a b b 6 3213 obbb 0 0 2 30 333 3 o A8 b 33 d>3o 18 7 “ 0 o b 3 2 l0 o 786 3 2 10o 789 “ 0 0 b 61 333 3o lib 3 3 6 5 o 120 “ 0 o 2 320 8 o08i 3z08o082 “OoO 32 3337 o033 tototoloOtol “0 o 3 3205o3b5 320 5 «.=• 362 0 0 3 to'3 333 d o V3 V 336b oV3V “Ool 3202 c626 3202 o62V “ 0 0 3 £> ^ 3360 o bZ^ 33o0 ob 2 6 “ 0 o 2 3 1 9 90 878 3 lv 9 o 8 6 3 “0 0 6 toto tototol o 6 V 1 33to2 ob96 “ 0 o 5 319 ? 0 118 3iV ? o l2 6 “0 o 7 3o 33bh otoHl 33 6 “s- o 65 2 “ 0 o n 319bo 352 319bo 353 “Ool 3 i 3366 o 39 a 33o6 o3V 1 Ool 3191o673 31V1 0 6b9 Oob tot 336B o li t ) 3 368 c 215 OoO 3188o7?2 3V 33? 0 o02 to stoliso 02 3 0 o 2 3 lb 5 0 963 3186 oVo2 0 c 1 HU 33?1oOiA .337lobib 0 06 toa8 3 o1 to! 3183 0 139 “0 0 2 H 1 3 3 1 3 o 398 3 3 1 3 o 393 0 ob 3180 0 30b HI 33? 3 o 33 7 3 3 7 3 o3 6 6 Ool 3 1 1 1 0b66 3177ob57 OoV b3 337? o102 3 l 7 b a 6 V? bb 337 8 o833 337 b a 83 d) Ool 317lo?27 3111 0 726 0 c 1 Hto 338 0 otobl 3 3 8 0 o65 i Oob 316 80 8 39 3168 o 8 bi “0 0 2 Hto totobioZbto 33b2a26b “0 o 0 316 6 o Vb3 2 16 5 0 9b6 “ 0 0 to H "i 33d3 o93b 3383oVb2 "0 o3 H b 3386 o 6 1 a 33t 5 0 6 lo “ Oob -X 205 SAbLE 36 o ubbE kV ED AiMU CALCULATED v) AV EHuMtJ E Rb I VAC 0 C M 5 0 1 „ 0 1 12 UF THe 0 0 1 0 1 - 0 0 0 0 1 bAND GF C 2 k 2 f f ooOLOeaaoCTq ■ cacacacai «=»«=» es» «s»«s» «a=«=» «sa «=»«. «> a k 4 J i 0 b S K 4 JD L A L C 0“C PI J 5 LdS PIJ JCALC (J“C A100 X1C0 i 3kVA o 006 3k.VA.aUV0 =>0 o4 2 3kV3o4 26 3k9 3 043A ■“00 6 3 2 8 lo 662 32 b 1 ob55 O o l 3 32 V5 o 76 h 329 5 0 765 “ 0 0 0 327 Vokbl 32 7 9 0 27 b 0 a 3 4 3k Vts o ObV 32V 8 0 091 “0 0 k 3k?6 0 901 327 6 0 89 4 • 0o? 3 3300 o4l0 3k?4o500 3274 c. 504 “0 0 3 6 330k o h 2 2. 327 k o 100 3272 0 10b “ 0 0 6 7 3305 oOdk 3306 o0k6 "OoA 326Vo702 6 3307 o.32 A 3307 0 32 3 “ 0 0 2 326 7 0 287 34b 7 0 291 ”0 0 4 V 330v ob16 330Vo613 0 o4 3k6 4ob71 3264c. 873 “ Ook aC 3 3 1 i o bV? 3311obV5 Oo k 326 ko 450 3262 0 4 44 Ool Al 33 l4o I Ik 3 314 0 17 0 0 o3 3 2 6 Oo018 326 0 o0 18 OoO Ik 3 3 1 6 o4 3 1 3 310 0 4 3 7 0c0 326 't 0 5 81 32 57 0 58 1 Ool lb 3 31 b o t> W 3 3 A 0 0 6 9 7 0 0 k 325 5 0 A3? 32 55 o 137 OoO 14 33k 0 o95A 3 320 oVhV 0 0 k 3232a 689 3252 0 6 8 7 Ook IE 33 k3 O194 3 3k 3 0 1Vh OoO 323 Oo22V 32 50a 2 30 “ 0 c 1 At. 33k6 o 4 3 A 3 3 k> 0 431 Ool 324 7 0 766 3 2 4 7 0 7 6 b “Ool 11 3327 t.660 3 3 k?o65V 0 c0 3 2 4 3 0 3C3 3 2 4 5 q298 Ool lb 33kVo.b7v 33k9 c.88 0 -Ool 3242o 829 3242 0 82 3 0 o? IV 333i_ o UVi. 333k oOVk “0 o0 324 0c. 344 3 2 4 0 q 340 0 0 4 k.0 333H okVE 3 3 3 4 0kV3 “Ool 323 7 0 8 32 3k37 o852 Ool 21 3 3 3 6 o h 8 V 3336 0 49 0 “Ool 323 5 0 3b3 3235 0 335 “ OoO k k 333 b o b7fc 3 33b 06 76 0 0 3 3k 3 k 0 8 4 b 3232 0 852 “ 0 0 5 k 3 3340 o b 3l 3 34 0 0 8 4 V Ook 323 0o 337 3k30 a 34 2 “ 0 0 5 k4 334 3 o0iA 3 343c014 “0 0 3 3227 0 823 z t 3345 o163 3 3 4 6 o 168 “Ook 3225okV5 3225 0 29? “ 0 a k 26 334 } O 30C 3347o3l0 -0o9 3 22 2 0 759 3222o?62 “ 0 0 4 21 3 3 4 V 0 43? 334V 0 44 0 “ 0 o3 322 Uo 2 A? 3 2 2 0 o 2 19 “ Ook kt 33b A ob5V 3331o 357 Ook 3217 0 6 6 ? 3217 0 665 Oo 1 2 v 3355 0 bbfc 3363 0 66 a “ 0 a- 3 3 k l 5 0 1C 1 3 2 15o1C2 -O oi 30 3355o?54 3355o Ii0 0 o4 3212o531 3212 0 52 6 Oo 3 3 1 335 7 obkV 3 3 5 7 c 8 k 3 0 0 5 320 VoV47 3k 0 9 0 9 4 2 0 a 4 3k 3 3 5 V 0 ObO 335 Vebb A Oofo 3207 0 345 33 33bA oV23 3 3 6 1 0Vk 1 0 0 2 3 2 0 4 0 ? 3 4 33b 3 0 VhO 33b3 0 V43 -Ook 3 20 e. 0 104 3 2 0 2 o 110 “ 0 o 5 3b 336 6 o Vho 330 6 o945 0 0 A 3 1V 9 0 4 7 0 3iVVo470 OoO 36 3367 cVk7 3 A V 60 824 3196 a 815 OoV 3? 335V 0 titl 350VoB87 “ OoO 3iV4oA43 31V40 A44 “0 0 A 30 3 3 7 1 00 k 1 33? A obk 5 “00 4 3191 0h55 3iV A o45 5 0 oC 3V 33 #30 ?35 3 3 ? 30 7 3 V “0 o4 3Ab8a?4b 40 3376 06 k6 337 6 06k 8 “ Ook 3 1 o6 oGa7 3lb6 D 0 k 1 “0 0 5 41 3 37 7 0 4 V k 337 7 o4V A Ool 3183ok 73 3183ok?4 “ 0 0 A 4k 337 Vo 333 3 3 7 Vc3k 8 0 o3 3180o 506 43 3 3 b 1 0 13V 3 3 8 l o 15 S Oak 317 7 a 715 3177 c.^16 “0 0 A 55bko9i4 338k 0VI? “0o,5 3 1 ? 4 0 V04 *>5 317 2 0 071 3172 0O68 0 0 3 HO 316 V 0 207 3 IbVo 2 0 b “Ool „ A 206 § A bi_E 3 7 UP 1 Ht 010(11)°+ “ 0000°0° bAlvD OF 12 C „ Z 2 2 a k 4 JDObb k (1 j 5CALC 0 “C P4 a 1) UBS HJICALC 0 “L x io o XI00 0 32B4 o2 46 3284 0 247 Ool I 32B6 o 383 32 86 0 38 4 Ool 3 2 7 9 0 349 3279o 549 0 oO 2 3 2b b o9l6 2.28 8 o9 13 0 0 4 327 S0 AB7 32 81 0 1 8 7 “0 oO 5 32VAo 233 32V io233 “0o0 3 2 14 0 8 A 6 32 7+o 8 317 “0cl 4 32V3 o&44 s2V 3 0 34>4 OoO 3 2 7 2 o438 32 72 o440 “00 2 5 32V3 o b46 32V3o 047 “ Ool 327 Co 049 3270o 054 “0a 4 b 32V8 o142 32V8 0 141 OoO 3 2 6 7 065V 3267 0 6 6 O “ 0oO 1 33C0o 423 3 300 0426 “0o3 326 5o 257 3265o257 0 oO b 330 2 o704 3302 o702 0 0 2 3 2 6 2 0 843 32o2 0 846 =■0 o4 V 3304 o9&9 5sQ4o Vo 6 Ool 3 2 6 0 o425 32 60 o427 “0 0 d 10 3307 o2 2v 3307 o223 “ 0 0 1 325 7o998 3257 0 999 “ 0 01 11 330V o 472 3 30V o4 7 2 OoO 3 25 bo 564 3255o 563 0 01 12 3311 o f k 3 3 1 1 0 70b 0o3 3253 0 1A5 b2S3o118 “ 00 d 13 33i3 oV37 33i3oV33 0 0 3 3 2 5 0 o663 3250ob63 “ 0oO 14 3316 o13 a 33l6oA60 Ool 324 bo 204 5246 0 200 0 o4 13 331bo336 33i6o355 Ool 324 bo 729 5245 o727 0 a d 16 3 3 2 0 o 34V 3320 o34b 0 0 A 324 3 0 244 3243 0 245 “ 00 i. 17 3322 o J 32 3 3 2 2 0 730 Ool 324 Go 756 3240o 753 0 0 3 i a 3324 o 900 3324o V0 A “Ool 323b 0253 3238o 251 0 0 2 IV 3327 o05 8 3327 c059 “0 0 2 323 5 c, 740 32 35 0 739 0 G A 20 33 2 V o204 332V 0 2 0 4 “0 c 0 5233o2lb 3 2 3 3 0217 0 0 A 2 a 3 3 3 Io33b 33b i 0 33 6 0 oO 3 2 3 0 0 6 b 3 3 2 3 0 0 6 64 “ 0 0 A 22 3 3 3 3 o 4 b 4 3333o 4 5 b “ 0 o3 322 bo 136 322b 0 140 “ 0 0^ 23 3333 05>63 3225o5b7 3225a 585 0 0 2 2 4 3337 0662 3337 0 6 6 b 0 04 322 3o 02i 322 3 0 020 0 cl 23 3 3 3 Vo73b 3 3 3 9 o73b 0 o0 322 0 0 4 4 4 3220o 442 0 0 d 26 5541 0 805 3 34 1 0 8 0 2> “OoO 3217 0 b 54 3<£ A7o854 0 0 1 2 ? 3343 0 634 5 3 4 3 065 5 “0 O A 5215 0 2 54 5 2 1 5 o 253 0 oO 26 334b 0 b bb 3 54 5 0 8 V 2 “ 0 0 3 3212 c633 32 12 0 641 “0G d 2 V 334 7 0 V0 8 3 347 cVI4 “ 0 c 6 3 2 1 Oo 013 3 2 l0o016 „0 0 A 30 3349 oV3L 3 3 34V0921 “O06 3 2 0 7 0 37 9 32 07 0 379 “ 0oO 3A 553 A 0 V A 4 3331 c,V A 3 0 c- A 3 2 0 4o731 3204 0 730 0 c A 32 3363 0 8 8 V 5333ob90 “Ool 520 2 0 067 52 02 oQ69 “ 0cl 33 3333 0 63 A 3-35 5 obD 1 “0 o0 319 Vo 894 3199 0 394 “ 0a 1 54 3 3 3 7 0 bO1 3 36?o 797 Oo4) 3 1 9 6 0 706 3196 0 707 “0 01 33 3 3 3 Vo?2 b 3" 33 Vo 726 0 0 2 3 1 V 4 0 00 7 31V4o007 0 oO 36 336 A o64 3 3 3 6 1 0 6 4 0 0 0 5 3 l91o293 3 1 9 1029+ - 0 a 1 3 / 336b 0 336 33b3o33 1/ “ 0 0 A 318 b 0564 318 8 c 567 - 0 0 4 Jb 5365 04 1 6 536 5 04 lb ” 0 o3 316 5o 827 31b 5 0 8 2 b “ 001 3 V 3 3 6 7 0 2b4 3367o2b3 Ool 3l83c,0?9 Bib 3 0 07 5 0 o3 40 336V 0 134 33bV 0 13 1 0 0 3 319 0o313 3180 0 309 0 0 3 42 3 3 7 0 0 V 6 3 5370a962 0 o3 3111a 533 31 $7 0 530 0 O ^ h Z 3372 077 7 33(/2o7 /6 Ool 317 4 o737 3 1 7 4 0 737 0 cO 43 3 3 i4 0 37 3 3 3 7 4 0373 0 c 2 317 A 0 9 2 9 3 a7 1oV30 “ 0cl 537 6 0 334 537bo333 Ool sAbVc105 5l69o 110 “ 00 4 45 337 b 0 i 14 537 bolls “OoO 3 i66o273 5l66o 275 - 0 Q d 46 337 Vo 633 337V oB5V “0 a 4 3 lb 3 0 4 2 6 31&3 O 42 ? “ 0O 1 4 ^ 5381 c.380 3381c585 “0 a 4 3 16 0 0 362 5160c, 564 “ 0O d 4 B 33B3o20y 33b 3 0 2 92 “ 00 3 3 15 (7 O 6 bV 315? 0 6 b? 0 O d 4-V 3 3 6 4 oV?V 5 3 8 4 o 98 0 “ 0 0 2 315 4o7v6 5154 0 795 0 Ol 30 3386 0 630 3386 o64V 0 0 2 3 1 5 1 obVO 3l5 lo b b b 0 0 3 3 1 336fco 301 3 3bbo 29 7 0 0 4 3 14b0 969 314H o 965 0 32 3 l4 6o02l 3 l4 6oC26 -0 0 “ 1 207 ’J AbLt 38 o UobttWfcD AhO CALCULATtD wAVtNUMBfcRS UVAC.CM J 1 1 0 12 OF lot 010(21) (II) - 0001 o e BAND OF C 2 e J M o U 0 b 8 R 1 *J } C A L C 0-C Pd^iUBS FCj DCa LC 0~C ;uoo JU00 1 3274o2j,3 32?4c2l4 Ool 2 327b o32V 32?b o336 "0oi 3264o 829 3 327 8.849 3278 oti!4V “ OoO 3 2 6 2 o450 3262 a 45 9 - 1.0 4 3281c. 133 3281ol53 OoO 326Oo 081 f> 32B3o450 3 2 B 3 o 447 0 o 3 3257o6B9 3257 o 693 “ 0 o 4 & 32bbol33 32 83 o730 0 o2 325 5 o 288 3 2 5 5 o295 -0.7 fl 3288 oC11 32 88 o004 Gob 325 2 o 8 8 ( 3232 o GbV G 3290.270 32V0 o 26 a 0 o 2 3 2 5 0 o 478 325Do 47 3 Oob 9 3292 o 317 32 v 2 o 322 “ Oob 3248 o 048 10 3294 o 76 3 3245ob21 3245 o613 o 0 tX 11 3296 o996 3 2 V 6 o W 7 “Ool 3243o 2.6 9 lc 32V9 o217 32V9 o2AV “0 o 2 3 2 4 0 o 715 13 3 301o428 3 3 0 1 o430 “Ool 323 0.253 323b o 2 52 0 o 2 A4* 3303o628 3 30 3 o 6 2 V “0o 1 323 bo 774 3235o 77V —0 o4 A5 3305.820 3303o817 0 o 3 323 30301 32 33 o296 Oob 10 3307 o992 3307 oVV4 “0 o2 323 0o 804 3230o803 0.1 11 3310 o161 33l0oA5v 0 o2 322 Bo 300 3228o300 -OoO 10 3312 o 3 24 3 312 o312 0 o 2 322 5o 791 3225 alo l Oob IV 3314 o437 3 3 14 o43 3 Oob 322302bV 3223 o263 Oob 20 3 3 A 6 o 5B2 33Abo38 1 Ool 3220o 731 3220o 730 0 o 2 2i 3 3 aOo 6 99 3318o6V ? 0 03 3 2 1 8 o 188 3218 o!85 0 o 4 <32 3320 o 7 95 3320 o 799 0 oO 3 2 1 5ob30 3215 o 630 0o0 2 3 3322 oiJVO 3322 o889 Ool 32 1 3o 0b2 3213 o 063 “Ool 24 2*324 o 9b2 .3 32 4 oV6 3 “0 o 3 32l0o482 3210.486 “0 o 4 23 332 7 o027 320 7 o 900 3207 o b97 0 o 3 2 b 3^>2Vo07 3 3 329 o 07b “0 o 3 320 5o 296 3205». 297 “Ool 2 1 3 3 3 loA1 a j> 33 a o A11 0o0 3202o 683 3202 o 686 “ 0 o 2 2 6 3333 o1.27 3 333 o13 I “0 c 4 3200.061 3200o 062 “ 0.1 £V 3333 o137 3 1 9 7 o424 3197 o 427 “0 a 3 30 3337 o A2 B 3 1 9 4 0780 3 1V4o7b0 “OoO 31 3 3 3 V o 104 3339 o103 “ Ool 3 1 9 2 o 119 3 I 9 2 o 120 “ 0.1 3/ 3341. 0&6 3 3 4 1 0Q66 OoO 3 lb Vo 445 318 9.448 “ Oo 3 33 33 v 3o 011 3 3 4 3 o0 a2 “ OoA 318bo7bl 3 1 8 6 o 764 “0 a 3 34 234 4oV4D 3 344 o942 0 o 4 3 1 8 4 0062 3 184o06? “Oob 33 334bo857 3 1 8 1 o 336 3 a 8 1o 357 “ 0 G 1 oo 334 Go 754 334b 07s6 —0 o 2 317 bo 632 3178o635 “0 o 3 31 3330 o t>4 1 3 360 ob3 V 0 o2 317 3 o ©99 3l75ob99 “ 0 o 0 3b 333 2 o3 04 3 352.507 —0 a 3 3 17 3o 133 3 1 7 3 o 151 0 o 2 39 33 34 o 3 5b 3354.358 “ 0 o 2 317 Do 391 31 7 0 o 389 0 o 2 40 333b o1V4 335boA93 Ool 316 ?ab!9 31b7 o 615 0 o 5 41 3358 oOlB 333b c0l2 0 c b 316 4o 82V 3lb4o827 0 o 2 4 2 33 3 V o 8 13 3 339 o bib “Qc2 316 2 o 027 3162 .02b Ool 43 3 3 6 1 o600 3 3 6 1 o60 2 “0 o 1 3I3Vo2l4 3l5 9o2il 0.3 4 4 3 156. 390 315b o 384 Oob 4 3 313 3o 543 3153 o 544 “ Ool 4 b 3 130o 688 3150.690 “ 0 o 2 4 7 3 l4 7 o b iv 3147 o 82 4 -0.5 „i 208 1 A b L c 39 o Obbc K\/t 0 AOLi C A L C u L A T ED 0 AV LNoMb tK i> dVACoCM ) uf The 010(21) 1f (II) - 0001 ^o^ bano of C M 2 2 J R 5 J 5 0 b S KSJDCAlC 0 “C P« J JOBS H i I CALC 0-C ;U 00 A100 -/\ 1 3274 o243 3 2 7 4 o 2 4 7 " 0 o 2 O 0 2 32?foo3b6 3 2 7 6 o 3 8 9 •=0 0 4 32b 4c810 3 2 6 4 o bl3 o 3 327b o923 32?b o925 "0 o 2 3 2 6 2 o 4 4 1 *9- 32bAo253 3 2 8 1 o 2 39 ” 0 o A 32b 0 o 0 b 1 32b0o062 “ OoO 3 32b 3 o 37b 3263 o5 7 7 “ OoO 32 57 o b75 6 32b3ob94 3^8 5 o892 0 o 2 323 3o2b8 32 3 5 o 28 3 0 o5 1 320b a 206 3 2 0 8 o20 A 0 o3 325 2o bb? 325 2 obS3 0 o 2 6 3290 O303 3 2VOo302 0 o3 325 0a4? b 3230 o4b0 “0 o 2 9 32V2 o796 324 8 o 060 10 329 bo080 329 3o082 - 0 o2 3245a 633 3245 o 650 0 a 3 Ai 32 97 o 3 3 9 3297o3b0 “Ocl 3243o226 12 3 2 9 V o 6 2 V 3,2.99 o63 (3 “0 o A 3240o 7b ? 3 2 4 0 o J 94 “Oob A3 3-3>0 AobbV 33CAoBVA “ 0 o 2 523bo35ti 32 38 o35b 0 o 2 A4 3304 o 19-4 323 6 o 9 il 3 2 3 5 o911 0 oi A3 3306 o 38b 3 30b o 38 8 Ool 3 23 3o 460 32 33 o458 0 o 2 Ab 330 b o 6 i 6 330b ob22 -Oob 3230c 995 3230 0 997 “ 0 c 2 A7 3 3 A 0 a b 4 7 322 b o 32 1 322 8 o 329 “0 o 2 lb 3313o0b3 33l3oOb2 Ool 322 ba 036 322bo053 0 o 3 IV 3 3 l 5 o i ? i 331 3 o266 0 o 5 3 2 2 3 o 56 9 20 3317 0^63 3 3 i 7 o 9 b 0 0 o 3 3 2 2 A o 0 7 6 2l 3 31Vo 643 3 3 1 9 o64 3 0 o 2 321 bo 377 3 2 i8o5?4 0 o 2 2 (!• 332 A o 8 19 3 32 A o 8 1 3 0 oi 321 bo 0b7 52 A 6 o 06 4 0 o 4 23 3 3 2 3 o97V 3 3 2 3 o973 0 04- 3 2 1 3o 343 3213 o 544 “ 0 o A 24 3326 o 1 2 A 3 326 o 12 3 Go A 321 A c 014 3 2 1 i o 0 A 3 “ 0 o A 2b 332b o 23b 352 8 o 259 “0 o0 320 bo4?5 5208 04?6 “ Ool 26 3330 o 3 80 3330 o33 2 “Ool 320 3o931 320 3 o 92 7 0 G 5 27 3332 o 4b 3 3 3 3 2 o491 “ Oob 320 3o 3ob 3203o368 OoO 2B 333d»o382 3339 o38 8 “ 0 ob 3 2 0 0 c 798 32 00 a-79 8 “ OoO 29 333 b o 664 3 3 3 6 o b 7 1 “ 0 o 6 319bo216 3l98o217 “ 0 o A 30 333b o7 40 333tio739 0 o0 3 A 9 3 o624 3A9 5c-b26 “ 0 c 2 Ai 534 0 o 7 94 3 840 o 794 0 o0 319 3o 022 53 93o022 “ Ool 32 3342 o 839 3342obJ3 Oob 319 0 a40 6 5190 o407 “0 a A 33 3 3 4 4 o B 5 8 3 Ab 7 o 779 3187o 791 “0 o 2 3*4 3 346 0868 318 5 o140 3A85o142 “ 0 o 2 3b 3348 o8b2 318 2 o 491 51 b 2 a 4 91 Ool 3b 33bQ ob4Q 3 a7 9 o827 3 a7 9 o827 0 a 0 3? 5 3 5 2 o b 02 317 7o 147 3 1 7 7 o150 “ 0 o 3 38 3 3 3 4 o 74b 317 4o 4o0 6 A 7 4 o 4 6 1 “ Ool 3V 333bo 6 bC 3 35b o67 8 0 o 2 3 1 7 l o 734 3 1 7 i o 759 “Oob 40 3b5b o 392 3338 o392 “0 o 0 316 9 o 046 3 1 6 9oC43 0 a 3 4 A 33b0o46b 33b0o489 “ 0 o2 3 lb 6 o 3 lb 3166 o 314 0 o 4 *12 3362 o372 3 3o2 o3 7 0 0 o2 3163a 574 3i 63o 572 0 a 1 J 3364 o25? 5 364 o 284 0 o 3 316 0 o 822 3 l 60 o b17 Oob 4J.4* 3366 o 078 3 3ti6o 03 2 "0 o4 3 1 3 boObO 313bo04b 0 a 2 43 313 5o 2 b6 3 1 55o 266 ■“OoO 313 2 o 468 3132 o471 —0 o 3 4 S 3 A49o 661 3 l4 9ofob3 “ 0 o 2 4b 3 A4bo 842 3146 o b42 - 0 a 0 _ A 209 U b L c 40 0 UDSfc k\f t L AMD CAL6ULA TEL) rtAtf EMuMB fcKb i VAL oCM > 12 Uh THE 010(12) 1 (II) - 00000! 1 bAMU Oh e e L2 n 2 J M J 5 Ubb K 5 OSCALC 0-C P4 8 ) UBS SCALC U-C A100 A A 00 X 32 76 o 779 32?6o?b0 —0 0 A X 327 Vo 10 A 3 2790102 -O ol 326 Jo3bb 32o7 0 Bbb “ 0 0 3 3 32 b A o 410 3 2b1o 414 —0 0 4 326 5o 007 3265c 015 “0 o? 4 3 2 8 3 o if A6 3283 07 A3 0 0 2 32b 2 0 62 8 32 62 o631 “ 0 0 2 3 32 bb o007 32bOo003 0 0 2 326 0 0 241 3260o 236 0 0 3 6 52bbo2v3 32b8o2fct4 0 o9 3237c 832 7 3290o349 3290 o333 “ 0 o4 323 3 0422 32 5 5 0 417 Oob a 32V2 o 806 3292 ob10 “ 0 o4 3 2 3 2 0 993 32 32 0 992 Ocl y 3xV3 o053 3293 o 037 “0.3 325 0o 3b1 3250 0 357 0 o4 10 329 7 o289 3 297 o 29 A “ 0 0 3 324 bo 117 3248 0112 0 a 6 3299c313 324 3o 653 324 5 0 6 56 i l 0 “ 0 0 3 12 3 3 0 l o 731 330x o 726 0 3 2 4 3 0 lb7 3243o189 “ 0 0 2 A5 3303o92b 3 2 4 0 0713 14 3306 o113 3 306 oil'} -O ol 3238 o225 A 5 330b o2fc7 J30do 290 -00 3 3235o 727 lb 33i0o433 33x0 o433 Ool 32 33 o2 18 i 1 3 3 1 2 o o04 3 312 ob04 “ 0 cl 3230 0 698 lb 3314 o 7 44 3 3 a 4 o 7 4 2 Ool 322 8 0 A6 A 3228 0 167 -O 06 i.y 32 ibobbb 3 3 A 6 o 8 6 7 —0 0 2 322 60 629 322 5 o626 0 0 3 20 331b oV80 3 31 & o 9 7 9 Ool 3 22 3 0 ii73 3 2 2 3 oG72 Ool 21 3 3 2 1 o077 332lo07b -O ol 322 0o 30b 3 2 2 0 0 308 0 o0 22 332 3 o163 3323 o 16J OoO 3/,i 70936 3217ov32 Oob 23 3 32 3 o 2 4 I 3 32 3c2 34 Oo i 3 2 1 5o 344 3215 0 344 “0 o0 24 332? O 2 92 3 321o2 92 -OoO 321 2 c. 742 32 i 2 0 7 4 3 “0 0 3 23 332 9o 339 3 32 9 o333 —0 0 1 3 210 0 184 3210 0 13 4 “0 o0 26. 3 3 3 A o 3 b 4 3 20 7 0 312 32 07 0 310 0 oi 2 ? 3333c 37 b 3 2 0 4 o 8 73 3 2 0 4 0 b7 5 “OoO 2 b 3 33 3 o37b 3333 o3 77 0 0 x 3202 0 223 3202 c227 “ 0 0 2 29 333 7 o 3o3 3337 o3b2 Ocl 319 9o 567 3199 0 367 OoO 30 333 yO 339 3 33 9 o531 0 G 5 3 l9 6 c 8 9 3 3A 9 60 b94 OoO 3 1 5 54 A o2 b 3 3194 0 209 32 3393 O220 3 3 4 3 o224 “0 0 3 3 1 9 10 308 3A9lo3lO “0 0 3 33 3 345 o 1 4 1 3 lb bo 79 i 31b6o7v9 “0 0 9 34 33* o if o 033 3 34 ?0OS4 “ Ool 3.18 6 0 01? 2 3186 0 07 5 —0 0 3 33 334b oV42 3348 o945 “0 0 4 31b3o337 3 b 3330 o bI 6 3 3 5 0 0821 “0o5 3 l 8 u c 3 e y 3i 8 0 0 38 7 0 0 2 37 3332 o b if 7 3352 oOBX -0 o4 3i77o b2b 3i77o 823 0 0 3 3« 3339 o323 3 3 5 4 q324 “0 o 2 31? 5 0 049 3173 0 043 0 0 4 3v 33 36 o 362 3356o33 2 l o 0 317 2 o 236 3172 0 2 35 0 0 2 h C 3 3 3 b o 169 3358 0 X63 0 0 6 3I69o44b Biby o430 “0 o 2 4 a 3339 0936 3359 0938 “ 0 0 2 3l6bo 6 2 b 3 1 06 a 6 33 -Oob 4 2 3361O 734 8361 0 737 ~0 g 3 3 l63o 807 3ic3 o 802 Oob 4 3 3160 0 937 4,4} 313 b d 0 3 315 8 0099 0 0 4 43 3155o227 4 6 313 2 c 338 5 X 5 2 0 3 4 3 “ 0 0 5 -A 210 TAbLE 41o OBSERVED and UALCULATED iai AVENUM6 ERS {VACoCM & O 1 12 1 UF 1 HE 010(12)Xf (II) O O O O M B A N D OF f C2 M 2 3 k 5 JJ O b S K 4 J J 6 ft L 0 0-C FtODUBS MJSCALC 0 “C X I 00 x io o A 327 bob A 2 327 6 od A 8 “ 0 a 6 2 327 9 o 161 3279ol62 -0o5 326 7 0 386 3267 0 377 0 o 8 3 j>2b A o30? 32blo50b -Ool 326 3o007 3265o005 0 0 i 4 3iB3 o845 3 2 83 o846 -0 oZ 3262 o623 32c 2 0 6 2 8 0 o0 5 3286 c180 3286c177 0 o2 326 0o 241 32b0o246 —0 0 5 t 3288 o 50 i 32 8 8 o 3 09 — 0 0 2 32 57 0 860 7 3290 c 8 2 4 3290 c.82 4 -OoO 3 25 5 0469 3235c.468 Ool B 3 2 9 3 o 1 3 / 3 2 9 3 o 1 3 9 - 0 o 2 3253o074 3253oQ71 0 ,3 9 3293 3295 o446 —0 o2 32 50 0 669 10 3 2 9 7 o7 4 3 3 2 9 7 o ?48 -Oo 3 3248o2&3 3248 0262 0 o 0 i A 3300 o 037 3500 c 0 4 2 - 0 o9 3 i 4 5 o 831 3 2 4 5 0 8 50 Oo A 12 3302 o32i» 3 302 o 32 9 —0 0 5 2 2 4 3 0431 A3 3 3 0 4 o 6 0 ? ■ 3 3 u 4 a O 0 8 —0 0 1 324 A o01 6 324A0007 OoV 1*9- 3306od7d 3 30b c 8 80 —0 a 2 323 8 0 579 3238 0 576 Oc-b* 15 3309 o 143 3309 oi^3 OoO 3236c-141 52 36 0 139 Do 2 Lb 3311o 400 3 311 o 39? 0 0 2 3233o 701 3i 33 0 695 0 06 It 2313 o 648 2 3a 3 o 6 4 2 O06 3 2 3 lo 243 32 3 1 0 2 44 “ 0 o 1 ib 331b obbO 3 3 A 5 c 8 7 8 0 0 2 322 be 782 32 2 8 0 7 8 6 “ 0 0 4 IV 3 3 Ab o 103 33i8oAU4 0 c 1 3226©320 2 0 o>320 o 3 29 3 320 c3lV OoO 322 3o 64? 322 3 0 847 Ool 2 A 33i 2 o326 3322 o52 5 0 0 3 322 A a 566 322 10 3b5 0 0 2 i 2 332 4 c 7 A 9 3326o 7 lb 0 o2 3 2 1 8 0 8 7 4 3218o 8 ?4 0 o0 23 3326 oVOA 3 32 b 0 898 0c3 3 i l 60 374 3216 0374 “ OoO i4 33iV o073 3 329o0o 1) 0 0 6 3 2 1 3o 856 32 A3o 865 “ Oob 23 333 a ©224 3 21 1 0 339 32A 10 346 “0 c? 20 3 333 0 36 8 3 2 0 8 0 813 3208 obI7 “0o3 2 6 3333 o'jVt) 3 333 0 69 8 0 o0 320oo 277 3206 0 277 OoO ib 3337 o fc A3 3 3 3 7 o 6 13 — 0 0 2 3203o 723 32 0 3 0 7 2 6 “ Ool 29 3 339 o?AJ 3201o162 3201 0 165 “ 0 a 3 30 334 A ©005 319bo 586 3 1 9 6 0591 “ 0 a 5 bl 3 3 4 3 0 878 3196o003 3A 96 ©006 “ 00 A 32 6343 o 930 3 54 3 c. 9 3 6 “ 0 06 3 a9 3 o 410 3 l 9 3 o409 0 0 2 33 5347 o9 S 4 339 7 0 97 8 •=0 01) 3 1 9 0o 799 3 1 9 00 799 0 o 0 3*9 3350 c 00? 3350 0 005 0 o2 318 bo 177 318 3 0 176 0 o 1 33 333 K 43 Dubi> K 4 jBLAlG 0 ~C P4 J 5 UdS P4 JDGALC 0 “L A100 X100 0 3 2 7 9 O970 3279©989 “ 2 0 0 1 3 2 8 2 o 3 U 32 52 o33 0 -1 »9 32 7 5 0 276 2 52 8 4 Q 662 3284o66 1 0 0 2 327 2 o894 32 72 0904 “1 a A 5 3250 » 97 4 3256 098 1 ”0 a 7 327 Do 516 3270 0 522 -Ocb ** 32S9o292 320 5 0 121 3268 0 128 “ 0 a 7 y 32V1o596 329 lo 393 0 0 3 326 bo 731 3 2 6 5 o 723 0 0 7 o 32V3o 6b7 5293 0 88 3 0 0 4 32b 5o 318 3263o308 1 oO 1 3290 o175 32 90 0 164 1 0 A 326 Co 893 3260c 880 1 0 3 o 3 2 0 2 0 12 OE J HE 0012 0 - 0002 0 bAND OE C H e e 2 2 K i J }Gbb K OJLALl Q-C P5 J) l)Bi> Pa j )calc U-L X100 X100 2 3z8Z » 7DO 3282o #29 2 o 1 3 328Do069 3zuDo0D6 lo 3 326 bo 6*52 3268 o 621 2 o 1 4 32b7o3v0 3 2 8 ? o37D 1 oD 326t>o248 32obo247 Ool D 3289 oC>93 3 2 8 Vo 6 80 0 o7 326 3o 864 32 6 3 o 8 6 7 “0 o 3 t> 3 2 VIo98a 3 2 9 1 o9d9 “ 0 o 0 3 / 6 1 o 4 8 1 1 329h-o 27o 3/94o 283 -0o7 32D9o089 6 32Vo oD7 3 3 z96 0D68 0 o 6 32D6o672 3/D6o691 -1 0.9 9 3298 o828 329boB43 - lo b 32D4o 2o6 32D4o 287 “ 2 o 1 10 3301oOao 3 3 0 a o108 “ 2 o 2 32Dlo 861 32D1o 8 7 6 “ A a 4 1 A 3303 o 3h 9 3303o363 “ 1 o3 3/4 9 o hhd 3249 c.4D8 “1 o 4 12 330D o D 96 330D oOOb - l o l 324 l i o 023 324? o 034 “ 1 oO 13 330 7 o 8 31 3 3 0 7 o83V “Oob 3244o 601 3244o 60 1 ”0 o 0 AH ^310 o 060 33i0oQ6Q “ OoO 3 /4 Z o 166 3 /4 2 o161 Oob Ab 3312 o 272 33l2o2f>9 0 o3 3 / 3 9 0 706 3239o7 12 “ 0 ob lo 3314 046 D 3237o 265 3 /3 7 o/D4 1 a 1 17 8316 ooD7 331006*5-9 0 o9 3234o 801 32 34 o 7 8 b lob la 3318 o 830 3318 ob18 i o 2 323 2 o 323 3232 o30 7 1.7 IV 3320 o992 3320o 97h lob 322 Vo 830 3229o 817 1 o 3 / 0 3323 o116 3 2 2 7 o32D 3/27 o 3 ID lol / A 33/Do242 322 4o 810 3 2 / 4 o 7 9 9 loO 22 3327 o 346 3 327 o3D4 “ 0 o 8 3 /2 Z o /7 1 32 /2 o2 70 OoO 2 3 332 9 oHDG 3329o h D0 0 o A 3 2 1 9o12D 32 l9 o 7 2 7 “0 o Z 24 3 3 3 1o 326 333lo329 “ Ool 3 2 1 7 o 161 32 A 7 o168 “ 0 o 7 zD 3333 o 5 8 V 3333oD93 “ 0o3 32l4oDb0 3 / l4o D94 “ 1 O 4 2 b 333 Do oh7 333Do6v O Oob 3 / 1 2 o 003 z / 3337 0-669 3209o396 2 d 3339 o682 3206 o 7t> 1 3 /0 6 o 772 “ 1 o i 2 V 3 3 4 1 o 6 ?1 3341 o o 7 8 “0o7 3204c. 1/1 3 2 0 4 o 132 “lol 30 33*5 3 o6D7 320 1o 468 32 01 o 476 -Oo'J a l 334D 0 6 12 33*5Do620 “ Oob 319 8 o b 07 3 1 9 8 o BOD 0 o 3 3*. 33h 7 o D#D 3 3h 7 o Dob 0 o 9 3196 o1/0 33 3193o423 319 0 o737 3a90 o717 2 o 6 3 d 3 AbOc.019 3a88 oOOD 1 O 4 30 318 Do 288 31 8 D o 291 -0 o 3 37 318 Zo 568 31b 2o5>81 “ 1 o 2 -I 213 I Ad l E 44 o GubEKVEu AMU LALCULAT ED irM9 EMUMb E RS ( VAC *CM ) 2 0 2 0 , 12 OF 1 He 0012 0 r - 0002 0 £ BAMU OF f fC 2 2 C^» csa «—». J r u } u b i> h4JHCAlC G—£ PS JflObb PIJ5CALC 0-C XI00 XI00 2 3262o750 3262 0 730 “ 0o& 3 32b5 o06 V 3265o0B0 “A o0 3268ob42 32 68 0 63b 0 0 7 *♦ 3 2 b 7 o 390 3267 o392 ”0 o2 32b b <>248 8266 0 250 ~0c2 5 32b9o 093 3 289ob9b “Ool 32b Jo 8b«i> 3 33 3 34 9 ob i 6O r—d 3 1 9 2 0 766 3A92o?93 -Oob 34 33bi 63b I o44b 0 3 19Go090 3l90o0b 4 O06 3b 3 1 b 1 0 364 3 1 8 ? 0 3b0 0 c4 3b 318 4 c 6 2 b 3l84o 6 2 1 Oob -1 214 TAbLE 45 o UbSEkvED AND CALCULATED w At/EimuMB ERb ( VAL.C* J OE THt 001(11)°+- 000(11)°+ bANU L.E 12 C ^ ^ — — ***““ rrm- «»-rr -as -r- a»««sx «*»«=_=»,o m n n o a*.. J K iJ) 4 Obi) KJ J 4 CALC 0-C P< O0UBS P5J{CALC 0-C X100 X1C0 O 0 338 3 ob^'V 3 2 8 3 o 6 l4 1 oO o 1 3285o955 3278c-902 3278o902 0 4 328b o28 7 8276o 531 3 3390ob07 3290 ob08 -Ool 3279-ol55 327^0151 0 o 4 4 3293 o 921 3 292 o919 0 o2 3271 a 764 3271o?62 0 o 2 5 3 3 9 5 o207 3 3 9 5 o2 IB - l o l 32b 9o 372 3269o 362 1 cj 0 6 3 2 9 7o BOb 329 7o 3 0 1 “ 0 o 1 3 2 6 6 o955 32b6o953 Ool 7 3399o779 3299o 784 "0 c 5 3264o 529 3264o534 —0 o 5 0 8302 oQ45 5 302 o 09- 9 “0 (?4 32b 2 o102 3262 © lQ 5 -0 o 3 9 3309 o 39b 8304o302 —0 o b 5259o 662 32 59 o 666 —0 o 3 10 3 3 0 6 o5^2 325 7o 221 3257 o216 0 a 5 11 3 308 o 77 0 3254o 749 3254o 7 55 —0 o 6 12 33l0o9bb 3 3 1 0 o9 b3 0 o 3 325 2o 280 3252 o264 “ 0 o 4 A 3 3313c £.93 3 3 a3 cA83 0 o 9 329 Vo 800 2249o801 -Ool 14 3315 o369 3247 o 306 3247 o 307 “0 a 1 13 3317 © 533 3317 c, 59 0 “0 a 5 8 2 4 4 o 808 3244 o 801 0 o 2 lb 33 a9 o705 3 319 0698 0 07 3242 o 2 84 1? 3 3 3 1 o899 3331 o b4 A Oob 3239 o 7 59 AB 332 3 o9b9 3 2 3 1 o214 3237 o 2 13 Ool 19 5 53 6 a 0 8 4 3234o 662 3234o66l 0 o 2 20 332b old5 323 2 o095 3232o097 “0 o 8 21 3330 ©277 3 330 o 3 7 3 0 o5 322 Vo 525 8229o 522 0 o 3 22 3332o 337 5332 o i9 8 - l o l 3 2 2 6o 936 3226® 936 -0 o 0 33 3339o909 3339 -0 a 9 3224 o 341 1 1- 3 + 3336 o 46 7 3 22 i o733 3 2 21 o7 37 o 0 35 353 b o 5 A 5 333b o513 0 © 2 3 2 1 9o125 KT 3b 53h0o55b 33 “9-0o353 0 o5 821 bo 511 3216o 507 o 0 37 3213 ob84 3b 3 2 1 lo256 3 2 llo 2 5 9 - 0 o 2 lAbi_E 46„ bbbE Ky E D A tot) CALCoLATEO wAVENUMBEKb JVALoCM ) 12 UF I Fit 001(11)°-“ 000(11)°- bANO OF C 2h 2 J R 4 j POdd R (J PCALL 0-C PC jPQfcS P 8d PCALC 0-C A100 XI GO 0 3279o725 i 32b2o07l 3282 oObd 0 03 32 7 6o012 327 5 0 014 “ 0 0 2 2 32fa*»o397 3284o4Gi “0 o3 32? 2o 643 3272 0 64b —0 0 2 3 3 2 bb o 729 3286o72b 0 o4 32 7 0 0 2 66 4} 32 59 o 04b 32b9oC«»0 0 o9 32bt 0 b79 6 32 9 A o 3 d i 329 10 34 3 0o6 3265o484 b 3 2 9 3 o 648 3 293 0 6 4 Q O 08 326 3o 076 32b 3 0 079 —0 0 0 1 329b oVIV 3 2 9 3oV 26' -0„7 32b Oo 6b2 3260o665 “0 0 3 d 32 9tt o 199 3 2 9 do202 -0 o 3 326 do2 37 3^bdo 24 2 -0 0 5 9 3300 o‘fb0 3 300 o46 d “O08 3256c 820 3265 odlO “OoO 10 3 302 07 2 4 3 2 6 3 0 3 64 3263 0 369 “ 0 0 b 11 3 304 0 969 3 2 b 0 o 920 32 6 0 0 919 Oo 1 lc 3307 0 203 3 24 8 04b 1 324 8 0 96 9 0 0 2 13 3309 0^2 A 3309o426 “0 o5 3245 0 989 3 3 1 1 o639 3311ob38 Oo A 3243*610 ^6 ■3313c 842 3o a 3 o 838 0 o*) 3t 4 X 0 0 2 1 ib 33 lb 0O2 V 331b o026 0 o3 323b ob22 17 33ifa 0 207 3 31 b 0202 0 . 5 323bo00b 32 36 0 012 —0 0 6 Ab 3320 0 3bi 3320o 366 “ 0 o*J 3 2 3 3 04V7 32 33 0 492 Oo 6 IV 3322 o bi 6 3230o9b5 3230o 9b2 0 0 3 20 3 3 2 4 06b3 322 6 0 422 322bo420 0 c 2 2 A 3326 0 7 76 322 6 0 867 32 2 5 0 867 “ OoO 22 3325 08 ?9 3328 c8B b “0 0 6 322 3 0 302 23 3330 0 979 3330 c 95 0 “Ool 32 20o72b a4* 3333 0 03>9 3 2 1 8 0 141 3218 0 137 Oob 26 3336 ol23 3215 0 636 ifc 3337 0 A 7 A 3 2 1 i . 920 32 i 2 0 921 “0 0 A 2 7 353 9 0 2 C3 3 3 3 9 0 203 Ool 3210 0 296 3 2 lOo29 3 0 0 2 2 d 33^10 220 3 3 4 i o 2 I 7 0 o3 320 7 0 643 3207 0 bb2 -Oo 9 2V 320 bo 001 3204o99fe O06 aO 320 2 0 323 3202 0 326 —0 0 2 -1 216 TablE 47o ObfaEKVEU ANO CALCULATED WAVENUMBEKb 4VAL.CM ) UF THt 001(ll)2e- 000(ll)2e BAND OF 12 C ^ ...... t t r ojoib,,, .. | c»«=>o»Q>aoo - ^ g_(J, CZ,__, *=«*-«-_~______i_-„____ J K 5 J i GdS RlJICALt o-c P 5 J Bubfa M J 5 C A L C U-C X100 X100 1 1 o 2 3 2 8 4 o282 3284c276 0 3 32b6 o600 327 Oo 1*»0 32 7 0 o 14-6 ■=*•0 ob 4 3288 o911 326 7 o 7b0 8267 o737 0 o 3 3 3291o2l4 326 6 o 361 32b5o 338 0 o4 6 j>cb>3ob0b 3 2 9 3 0508 Ool 326 2 c 9**9 3262 o 949 -Ool 7 3 2 9 3 o 792 326 Oo 524 52b0o 532 -Oob fa 3 29 8 oObb 323 8 o A08 3258 o103 0 o3 V 33 0 0 oJ3V 3300 a 336 0 o3 3 2 5 5 06 7 1 3^ 3 5 o 670 Ool iO 3302 obVO 33U2 a 396 “ 0 ob 323 3o 22b fa2 3 3 o 2 2 6 Ool 11 3304 o fa57 330a5>o 848 0 o 9 3250o 771 3 2 5 0 o 774 —0 o 2 b 3 0 ? Ooee 3 3 0 7 o093 -0o-5 3248o 31b 3248 o 314 0 c. 3 13 330V o. 32 7 3 30V o 330 “0 o 3 3245 0 8 4 b 14 33 A1 o 33 V 331 A o36i —0 o 2 3243o 37A 3243 o 3 71 0 o 3 13 b 3 a ^ o ?B3 3313o7fa3 OoO 3240 o fa9u 0 O lb 3316 o C02 331bo003 ■=■0 c 1 323 fa 02 5238 o401 c 17 33lfao207 33 ifa o 2 ib —0 o 8 3233 o907 lb 3320o420 3320o^22 —0 o 3 3233o 402 3233o*»0fo — 0 o. 3 19 332 2 o b33 3 322 ob2 3 1 o2 323 0ob99 32 30o 900 “ Ool 20 3324o b27 332 4-o 81V Go fa 322 8 O390 3228 o389 Ool 21 3 3 2 7 o009 322 3ob6 7 3 UF 1 Ht 0 0 1 (ll)2f - 0 0 0 (ll)2f b A N D OF 1 2 ° 2 H 2 J R I J >udi» rtlJILALl 0-C RSJ^libb F J J ) C A LC U~C X I 00 Xl 0 0 X 3xd4o282 32b40X 76 0 o7 3 3 2 b b o 3 9 7 3x7 0o 140 3 x 7 0 o 145 — 0 0 4 4 32bb o90b 326 7o 7o0 3x 67 0 7 6 6 0 o3 6 329lo210 326 6 o 361 32fo6o358 0 0 3 O 6 3x9 3 a 5 0 8 3 2 v 3 o30A 0 326 2o 949 3x62 0 95 1 “ 0 0 2 7 3296 o 763 32b Oo 624 JX 6 O 0 634 - 1 o0 b 3X9bo 056 3298o055 Ool 326 bo 108 32 5bol0b “ 0 o 0 9 3 3 0 0 o 3 1 6 325 5© 6 ?1 32 65 o673 “ 0 0 2 1 0 3 602 o6b7 3 2 6 3 o 2 x b 3x63o2xb “0 0 A 1 1 3 3 0 4 o 80V 3260o771 3250 0 774 “ 0 0 3 X bF THE 0010°2° - 0000°2° b AN D OF 12 C H 2 2 j KU»Ctib K 4J iCALL 0—C P 4 jiJOtJb P 4 J5 CALC 0-C A100 xioo 0 3cbOa303 3 2 8 0 o304 “ 0 o 1 A 3 l b lo649 3262 ob^l 0 o 2 3275 0 589 2 3 UF 7 HE 0010°22 - 0000°22 bANU OF 12 i n e e 1 1 — —— ■= «S» «—«t. a.Q><=>0 . J k ij)Ubo K 5 J A CALC U-C PC j SCjBO FCo JCA l C U“C X J 00 X100 2 32b3o 0^3 0 2 8 3o 0*$^ “Ool 3 3 c t l o 3 6 1 3 2 b 7 o 3 b 8 - 0 . 7 3270o 903 ** 3 2 b 9 o 6 8 3 3268 o 613 A, 3 2 9 l o 9 8 9 3260o120 3200o113 0»7 b 3 299 026 7 320 3 o 709 32 0 3o 709 “ 0 o 1 7 3 2 90 o376 3 2 0 lo 283 3261c 286 “ 0 o2 t 3298 o 830 3 298 o S 5 7 Ool 325 8 o8 0 i 32 :>8 0 8 b 0 0 q 2 9 3 3 0 loi2A 3 3 0 1 o 130 -OoV 3256c929 3236 o 929 0 o 5 1 0 3303o 393 3 303 o 39 0 — 0 o 3 325 3 o 9 7 7 3«:63o980 “ 0 o 3 ii 33 0 3 o o 5 9 3 303 o63 4 “ 0 o 1 3 25 1 o 52 b 32 5 1 o 5 28 - 0 , 0 1 2 3307 o V11 3307 o 90 O 0 o 3 3299 o069 l J 33i.iic.iol 3 o i 0 o i 3 2 io 0 329 6 o 603 32^6 o 6 0 2 0 o 2 3312o3b8 3 312 o391 “0 o 3 3299 o 128 13 33 o 6 2 *) 3 3 A 9 0625 “ Ool 3 2 9 lo 635 32 9lo 697 OoV lo 331o o Boo 3 316 o 833 1 o 2 3 2 3 9 c 101 o 2 ? 9 o 160 Ool L~t 3 3 1V o 08 jl 3 3 l 9 o077 0 oA 323 0 o 6 6 A 32 3bo 6 6 'if “ 0 o 1 l a 3321 c.2b6 3321 o296 “ 0 o 9 323 9c 168 32 39 o A 6 V -lol IV 33,2.3 o 3 !<£. O323o310 Ool 323 io 039 3 2 3 1 o 6 6 ? -0o7 2 0 3320 o72 1 3229o153 3229o169 — 0 o 5 2 1 332 1 a 929 3 32 7 o 92 8 Ool 3226o69l 322o o698 “ 0 o o cl 3330 o120 3 330 o132 “lol 322 9o192 322 9o 139 0 o 8 c3 3332 o33 A 3332o332 0 cO 3 2 2 lo613 3 2 2 1 o 6 16 -Oo 3 2 *? 3 33 9 052 9 3 2 1 9 0 101 3219 c096 0 o5 c o 333o o 7 it 3 3 3 0 o 723 “ 0 c.3 3 2 1 Oo 385 32 i6 o 573 lol 26 333 b o 918 3 3 3 B o 9 13 0 o 9 3 2 19 d 039 3219 o09 V 0 c9 cl o 3 h 1 o 1 0 a 3 39 1 o 10 3 0 o 2 3 2 1 lo3l 1 3 2 1 lo52 3 -Oo 7 C b 3208 o 996 ,a 9 3206o 965 3 2 0 6 o969 “ 0 ,3 -x 220 T a bl£ 51a LiCicKVtl) AND CALCULATED WAVtNoMbcKis IVAC oCM ) UP 1 HE 0010°22 - 0000°22f BAND Oh 12 H ^ — «=, «=-- . a » s i , C B . e9s»n-i n, rm T. m. e» «n> ce»«w «b J f t 4 J i Gbi> M O B t A l C 0-C P 1J J Ubb P 4 J ) C A LC 0-C A 1 00 X100 2 328300^3 3283oG50 “0 o? 3 3287 o 361 3287o3?3 “ 1 o2 32 70 o 907 H 32b9obb? 3268o 517 3 3 2 9 1 O990 32b 6 o120 3 2 6 6 q 117 0 o 3 6 3294 o284 32b 3o 704 32b3o708 — 0 o 3 7 329b o36 8 3 2 6 Ao283 3 2 6 1 o 290 -0 o5 t> 3 2 9 8 obbb 3 2 Vbo842 lob 323 8 o 861 325 8 o b62 -OoO V 330 ia a.2 i 3 3 0 1 o10 3 lob 3236o429 323bo423 0 o 4- 10 3303 o 349 3303o 33b —0 o9 3253o 97 b 3di53Q97 8 OoO 11 230bo 596 3305c 601 -0c,6 325 Io 328 3251o 522 Oob A 2 3307 o b31 3307 c.b33 "“0 o 2 324 9c 032 3 2 4 9 o 037 “0 o 4 13 33 lO oOfcli 3310o034 Oob 3246o37b 3246 o582 —0 ob A4 3312 o 27 2 3 3A.2 o2fe 3 Ool 3 2 44 o 101 3244 o 09b 0 o3 A3 3314 o4b3 3 2 4 lo 603 3 2 4 1 ob04 0 © 2 lb 3316 o63? 331bobt>3 0 O 4 32 3 V o100 1 i a 3 a 8 o 830 33A8 o 8 30 “0 o A 3 2 3 6 0b87 a b 3 3 2 0 o992 3 3 2 0 &99b -0o5 3 23 4o ObO 32 34 o06 3 —0 o 4 IV 3323 olbl 3 2 3 1 o 532 40 3323o291 3 3 2 3 o 293 “ 0 o2 322 bo9b8 32 2 8 o 990 —0 o 2 2l 3 3 2 7 o <■} 2 7 332? o424 0 o3 3 2 2 b o 44 0 3226o‘!o3b 0 a2 44 332 9 ob^» 1 3329o642 “Ool 322 3o b bb 3 2 2 3 o 876 0 ob 4$ 3331o 638 3 3 3 1 o 6 49 “ 1 o 0 322 Io 304 44 3 3 3 3 o 7 4 a 3333o?43 “0 o2 32lbo7l9 3218 c 722 “ 0 o 3 4t> 3333 o826 3333a 824 0 3216 o130 Uf- InE 010(31)?+ - 0002°0° BAND Uf- 12 C 2h ," s»a •» o» a» c» n o «—«* ™rr** “**“ K 5 { u B b K { D C A L C U“C (■>«31Ub8 PS J ) CALC 0-C XA0 0 /a oo 0 3260 eO l3 1 3Zfc3 o 133 3 Z o 3 o 158 =0 ob 3^.b 6 o 097 Z 3 ,2 6 b o494 3253a 72b 3 3z 6 7 obOz 3 2 o 7 o 8 2 2 ““2 o 0 3 2 b lo 339 3 2 b 1 o 3 9 9 — 0 o 5 9 3Z 70 o AAO 327 0oA42 - 0 o 2 32-5 8 o9 59 3 3 z7 2 o959 6 z9 6 o b b b o 327^0 /bo AzI4) o 7b V “ 0 o 2 32*59 o 1 9 / t 32? 7 o O b1 3277o0b6 “ Oob 3 2“} 1 o 749 3 2 9 1 o 731 1 © 8 b 3 2 7 Vo3^0 3 z 7 V o 3 9 7 “ 0 ©7 3z39o 318 3z39o 3 06 1 o 2 V 3zb A o63b 3Z b lo630 Oob 3 z 3 o o 88 1 32 36 o 8 ? 2 lo b AO 3 z 8 3 o 9 1 b 3283o906 A o 0 3z34o436 3z39 o931 0 o 6 i i 3 z 8 6 o 17 9 3 Z o 1 o982 1Z 3Zbbocib6 AZbbo“)40 1 o Z 322 Vo bZb 82Z9o 326 “ O ol 13 3z9 0 o 70V 3290a?04 Oob 3 22 7 o 0b7 3Z27o062 -Oob A9 3z V z o V b V 3292 o9b 7 0 o2 3 2 2 4o bo 3 3z29o 592 “ 0 a 9 Ab 3z9b <>207 3 z 9 5 c z 0 ^ 0 o 3 322 2 o10& 3 2 2 2 o 1 1 6 — 1 o 0 At. 32V? o4bl 3297 0 o 6 32A 9o 621 3219 a 639 -1 «3 A 7 32 VV o 6?9 3 299 ob 79 -Oob 3 2 1 7 o 139 3 z 17 o 196 — 0 o if Ab 3 3 0 1 oVOb 32l-5o 6b9 3 2 l9o 633 Oob IV 3 3 0 “} o 1 2 6 3 2 1 2 o 198 3 z 12 o 139 “ 0 o 6 zO 3 3 0 6 0338 3209o6b2 3 2 0 9o 69 9 0 o 2 2 A 330b o b 3 1 3 3 0 b o39 0 — 0 o 9 320 7 o14 A 320 7o 190 0 o A ZZ 33 a 0o 73b 3 3 A 0 a 7 3 4 0 o A 3209a 62? 32 09 c 629 0 o 3 23 3 3 lZ o 9 l7 320 2 o 109 3Z02 o 103 0 o A 2 *» 331b oU56 3 3 ib o 0 b 9 “ Oo/ 3 AV9 o 376 zb 3317 ©z36 3 3 i 7 o 2 “j V “ 1 o 3 3 A9 7o 096 3197 o09 2 0 o 9 26 3 3 1 9 o397 31V 4c50b 3199a 301 0 □ 7 2 7 3 3 z 1 o b32 3 1 9 lo 9 b l 31V1o9b2 “Ool ZB 332 3 o 6 0 A 3 3 z 3 o 6 b 4 1 c0 3 A8 9a 397 318 9 o 3 99 0 a 3 ZV 33Zb o76 / 3 3 2 bo 7b2 Oob 318 6 o 829 b lb b a O z b “0 a 4 3U 33z 7 o bbo 3 3z 7 o b b ? “ 0o A 31ti 9o 2 bb 3 l 89 o2 31 0 0*5 31 33zVo993 3 32 V o 940 0 o3 318 A o 63b 3 1 8 1 0663 “ Oo / 3Z 333Z oO AO 3 3 3 za 011 “ Ool 317 9o 069 33 3 3 3 ^ 0 OoV 3334o073 “ 0 o 9 31 7 6 o9bb 3A76o952 0o-» 39 317 3o820 31 73 o8 2 7 “ 0 o 6 3b 3 17 1 o 193 3171ol88 Gob 3o 3 1 6 0 ob9l 316 0 o 33b Oob 37 3 16 b o 8 61 3 l 6 5 o 8 7 i -1 o0 3b 316 3©193 3 V 3 16 Oo bOb 3 1 6 0 © 309 0 a 9 - 1 222 TABLE 53 o ObbEkVED AND LALCULAI ED WAVENUMdEKb IVACoOl ) 2 0 12 UF Iht 010(31)2 (II) - 0002 0 BAND OF e e L 2H 2 J k i J lUbi> k t J 1) CALL 0-L FIJ JUtiS FI J }L ALC U~C XI 00 X100 2 3z64 <>34 a 3264-o 330 I o2 d 32bo o 66 b 3206 06 3V OoV 3250 o211 4 3zoa o v90 d2obo980 Ag O 324?o bio d24? o 838 “ 0 o 2 d 3z7io 292 327 A o 292 -OoO 3 24 5 o 4 5 7 3245 o458 “ Ool 6 3273 o3V5 324 3 0 0b0 3243o 073 ” 1 o 3 1 3z75o867 327 3 o B 8 8 — 0 o A 32 4 Co 677 3240 o 6 82 — 0 o 6 a • 327 B o170 323bo285 9 d Z 80 04d4 3280 o441 “0 a 7 3z35o881 10 3zb2 o699 32b2o?00 “*0 oO 323 b o 460 dz3 3 o4 6 9 “ 1 a0 11 3zb4o 9db 3 2 84 0 996 -OoO 323 1 o 0 4 4 3 2 3 1 o 050 “ 0 oO 1 2 32 b7 o175 3 z 8 7 o A 7 V “ 0 o4 322 b o 623 3228 o 622 0 o 4 13 32b9o399 3zb9 o400 “ Ool 3226 o187 3226o 185 0 o 3 j.4 3z 9 1 c 59b 329 A o 60 7 ” A o 2 3223o7*»8 3223a 7 38 lol A I> d2V3 o bOd 3 2 9 3 o b O 1 Oo2 3 2 2 lo 2 89 32 2 1 o 280 0 o 9 10 3295 o9b8 32V3o9dZ 0 a 6 3218 o 811 17 325e o A4Z 3298 o149 ”0 o 7 3 2 a 6 o 3 3 4 32 16 540 3185a 537 0 o3 dO 318 2 o 893 3182 o890 0 o 3 31 3 18 Go 2 34 3 IbOo233 0 o 2 d Z 3 il l o564 33 31740 btiV 3i74obb5 0 o4 34 317 2 o190 3172 o194 “O o 3 35 316 9 o 480 31b9o4 89 “ laO 36 316 6 o 778 31o6o ?70 Oct 37 3i64oG31 31b*»o 033 “ 0 o 2 1 223 T A t>Lt: 54 o Cbto cKVtb AML) CALCOLATc.0 in) A V EMUMb EK i> 4 VAC.Cm } 2 0 Ob 1 ML 010(31)(II) -0002 0 b AM D Of 12 C H 2 2 vO 8 1 } ObS KJjiSLAlC 0-C P5 JDGbS P 4J ) C A lC 0-C XI00 XI 00 2 3264 c3^I 3 2 6 4 034? “ 0 o to 3 32toC o toto to 3206 o to? 3 - 0 o? '32 bG O 2 2 1 4 32tobo990 3268 09 93 -Oob 324? o 83to 3 2 4 7 ob4 0 “ 0 o4 to 327 Ao292 3 2 1 1o 30 7 “ 1 o 3 3 24 bo 4b? 324b o452 Oob to 3273 o to 11 3243 o OtoO 3243 o 0b7 0 a 4 7 3273 o90to 3240o 6b4 a 3 d l b a lb 9 3 2 ? 8 o l9 4 “ 0 o to 323bo2b3 323bo 243 i e0 9 32b0 o4?4 32b0 o4?3 0 o A 323b o b2b io 32 82 o730 3282o 7 4 4 0 1)6 3 2 3 3a 402 3 2 3 3 o 4 0 0 0 o 2 n 32 63 o 0 0 1 32b b o00to 0 o 2 323 0o Vob 3 2 3 0 o 966 -0 o 2 id 328? o 26b 32to(7o23b Oob 3 22 8 o b27 3228 ob 2 b Ool o 3 2 26 o 0?? 13 32 b9 o b04 3 2 S 9 302 0 o2 o I4? 3 2 9 1 o 7 3V 3 2 9 1 c 736 0 o 3 322 3o 622 322 3 o 62 0 o A3 32 V3oVo^ 3293c9to 0 o 3 32210 ibb 16 3 £.96 o 1"? 3 329e> ol 76 “ 0 o2 321 too 684 3 2 1 8 o too2 0 o 2 A? 32ybo3ttti 329b c38l 0 , 1 32lfco202 3216 o 201 Ool A 0 3 3 CO o 373 3300o37to “ Oo 3 3 213 o ?09 3213 c 71 1 “"0 o k 19 3302 o 7 toO 3302 o7to0 -OoO 3 2 llo 2 2 3 32 1 1 o 213 lo l 2 O 3 304 o934 320 b 3 3 1 3cto2 b 319 to o C 31 26 3317 o7 2 b 3 3 I I a S 3 A “0 o 2 319 3 o 4 60 3 A 9 3 o 4 6 / - 0 o 1 2 7 33 A9 c.to2 0 3l9Gc8bto 3 ^.90 o 693 "Ood 3 3 2 lo 893 332 lo8V3 0 o0 318 8 o304 3188 c 308 “O 29 3 3 2 3 093 i 3l8b>° 714 3ib bo712 Ook 3 0 332 6 o DC4? 3 326 o003 -0o A 3183a10b 3 A 3328 o038 318 0o491 31b0o48? 0 o* 32 3 3 3 0 oOtoZ 3330o0bto O o b 3 17 7 obb4 3177 obb? “0 g 3 33 3332 oOtoV 3 A? bo223 31 lj o 2 i 5 Oob ! £ o 34 3334o 04 1 3334 o04to o 317 2 o b ? 3 3 a 72 o btoO 1 o 3 33 3 1 6 9 o 892 31to9o 893 “Ool 3o 3 ito $ » 204 316? o213 -OoB T AtiLE 55 * DESERVED AND LALCULa TED WAVENuMBEks ( VAC «LM ) DF THt 010(22)°- - 000(11)°- BAND OF 12 C ^ 2'U>0> U U ’U U JV U u Uju, tv U U'(a-0 • ijg U> 0, (a W U>< U. (aU U (a-Up W W W s J X I 8 O { IjjUi U' 0- OJ WU>UJOJO-'UJU>UJU;U,U'0>UJ0.’UJU1U CsJ W WUj IaiU>UJU!IsJWUJC(.,U>WU»(jJ1jJWIsJWU>U>U.,U. WUUtOJEV EV) 5 «=> 3 } O c p vCv:>{ vj.>f ifdffoaaccoi'j^'-j^aff R-e e s - w - t f ^ u.|^U)ijJWNF\w\-w!»i-'t»HOooOv;"i j c 5 9 ° 0 AfxiL) 8 CE*4V 4 f\ O yJ 4 V <5 OCT -4VU- ^ V? =»’P •? P -S' IN. O E P .? t= O ^ JU - UJt-'vDC ^ r v O ^ U U-t-’CnO ^ c=->£-»J 3 =■ ! j j 000 OQOOOOOOOOOOOOOOOQOOOOOOOOOOOOOOOOOOOQOOaoOQOoOOO j {- 5 i I i -*J c=-U‘CXFV'U'CBC--W0rX>OfVlv O'"=JvCOt~C>-(V(\.g;U’EVtvK't-'OvS0PU,OJN'OCPW C-'Cr0U'C-^,-S t>C V W 3 !> 9 ) iJ CALLULATfcL) A V fc N UM B t K iVACaCM i> 8 u'VJ Wi ^ c-0ON»j-'6-'Cd4> «=j»£:vf a-^vtfs-OJfMvCu © o o a ^ou.'J'■>cs,>5^'Ocot-'0” 0 “>Jwcofvu =o«=j«»J0 vjiPv,ca j r ! 6 ^ 0 Ot^.’ *=J J? ^ - a u P C < 2 C O f\a,0 ^ O 0'P' O O f ^ U 'C ’O"-? U«»,iV'OC“ U'^-0 UJa^CtCCP MU'U-U ,5 vCvTON-'J' 3 r- 8 0 S E 3 ! t o 0 8 8 i o E 8 f 8 8 ! 8 5 3 i 8 8 8 8 8 E E C 8 8 8 S ! I i 8 E 5 8 8 >: 8 8 ° E OOO OOOOOOOOOOOOOOOOOOOOOOOOCOOOOOO QOOOOOOOOOOOOOOOO 8 f-'O 8 8 ° 8 oooooooooooopooooooooooooooaoooooo ooooooooooop° ooooOO!8 8 8 Q-G- p-> rvVJ’C.COt-'C— fV'OOEVU't- ^ Of- 0 ■$■ [—1~ t-lOi-JONiWWi>'bOH c- Ht-'Ofe t- O ON ^‘*“Off fVt --0 OO | j f 6 8 5 1 8 8 { f I 8 8 5 3 8 E i> u*u'u. a'u'u>uu^w ^t^ou'u'uuw w u>'uw ^u’u'(jjui 0pg,ujuujoou,o266 0404 “ Ool V 330B o449 33Qdo4bO “OoO 326 3o Vb1 32b3 0979 0 0 3 10 3310 o 695 3 3 10o bV 1 0o4 3 26 1 0 542 3261 0 543 “Ool £ A 3312 o9 23 3 3 A 2 o 9 2 1 0 0 4 325 9oOVb 32 b 9 0 0 9 8 “0 0 3 12 3 3 1 3 o143 3 31 b c, 14 0 0 0 4 32 5b 0642 32 5b 0 642 “OoO 13 3317 o331 3 317 o34 b 0 0 3 3 2 b 4 o 180 32b4o177 0 03 A4 3 31V o 344 £>3a 9 o 54 3 0 0 2 32b Ao 705 3251 0 70 1 0 c4 13 3321o 730 3321o72b 0 o4 324 9o 220 3 2 4 9 0 21 4 Oob Ab 3323oV00 3 323o 8 9b 0 0 3 324bo 719 3246o 717 0 0 2 A f 332b o037 ^32 6 o0b4 0 o3 3 2 4 4 0 2lib 3244 0 209 0 0 7 as 3326 o1VB 3 328 o198 “OaO 324 Iofe91 3 241 0 689 0 0 2 IV 3330 o 3 30 3330 o32 V 0 0 1 323 V 0 1b 1 3 2 3 9 0 15 6 0 0 3 <2 0 3332 o44C> £>332 o44b OoO 323 6 0 61 b <£. 1 3334 o 34 9 3334ob4b Ool 3 2 3 4>o ObO 3234o0b0 “OoO ^2 3336 ob37 333b o b 3b Oo 1 3 2 3 lo493 2 3 333 bo 713 3 3 3 8 o 70V 0 0 3 322 8 0 V a6 3228 0 vl4 0 0 3 i ■A 3340 o 770 3340 o 7b? 0 0 2 322b 0 321 23 3342 o 80V 3 34 2oB10 “Ool 3223 o?lb 20 £>344 o 8 £>3 3 3 4 4 o b3 b “0o3 3^21 oOVB 2 7 3346 0844 33*o6 & 847 “ 0o3 3218 0 467 3218 0 466 0 c2 2b 334b oB37 3348 o84l “0 0 4 32lboU22 321bo 620 0 0 2 2 9 3330 oBlb 33b0o819 “ 0 0 3 3 2 1 3 o 157 3 2 1 3 o 160 “0 0 3 30 33b do 3 Si 33b2o?80 “0 0 3 3 2 1 0o482 3210 o48 7 “0 0 4 31 33b4 o72 3 3 3 5 4 o724 Ool 320 l o l l 8 3207o 79V “ 0 0 1 32 33b b o 0 47 3 3 5 b ob50 “0 o4 320 5o092 £>205 oOVb “0 0 4 66 33b8 obbv 3358o 560 “ Ool 3202 0 37b 3202 o3b0 “ Oob 3h 33bQ o 4b1 33b 0 o4b 2 “Ool 319 9 0 643 319 9 0 6 4 8 “Oob 33 33b2o 328 3362 o 327 Ool 31960095 3196 0 9 02 --O06 3b 3 3 b 4 0 A 78 33b4ol85 “0 0 b 31V4ol43 3 194 o 142 0 0 2 b7 3366 o02 b 319105b2 3l9l 0 366 “0 c-4 3b 33b7 obsO 3 30 7o b4b 0 o2 3 Ab8 0 377 39 33b9 o 652 3 3b9 06 b 4 “Ool 316 5 o??4 3185 0 ?7 2 0 0 2 40 3 371o 44 b 3 3 71 c> 4 4 3 0 o3 318^o 955 318 2 0 954 — 0 0 A 41 337 3 o 219 3 3 7 3 o2lb 0 0 4 3 1 8 O o 128 3ib0 0 121 Oob 42 3 3 7 4 o V 17 33? 4o 9 ? 1 0 0 b 3i7?o 28b 3177 a 27b 1.0 48 337 b o 716 33?bo7l2 0 0 4 317 4 0 423 31?4o4lb 0 0 8 44 33 ? 8 a 44 1 337bo*5 3? 0 0 4 3 i 7 1 0 543 “43 33B0o 1*41 3 3 d 0 o 14 7 -Ool 316 8 0 bbO 31b8 0 b58 0 0 2 4b 338 A o 840 3 3 8 1 o844 — 0 o4 316 bo 759 31bbo 761 — 0 0 2 4? 3 lb 2 a B4b 3162 © 8 53 “ 0 0 6 A 203 SAdLE 34 a UbSEKVEu AND CALCULATED wAVENUMb Et<6 I VAC oCM«s»eaoocaoaacaoo <=-«t» Co eo an «=*»