Citation
UV reflectance of frosts composed of water and ammonia

Material Information

Title:
UV reflectance of frosts composed of water and ammonia
Alternate title:
Reflectance of frosts composed of water and ammonia
Creator:
Pipes, John Gilbert, 1945-
Publication Date:
Language:
English
Physical Description:
viii, 115 leaves. : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Ammonia ( jstor )
Annealing ( jstor )
Calibration ( jstor )
Frost ( jstor )
Lighting ( jstor )
Monochromators ( jstor )
Reflectance ( jstor )
Solids ( jstor )
Vapor pressure ( jstor )
Wavelengths ( jstor )
Aerospace Engineering thesis Ph. D
Dissertations, Academic -- Aerospace Engineering -- UF
Frost ( lcsh )
Planets -- Atmospheres ( lcsh )
Ultraviolet spectra ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 112-114.
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by John Gilbert Pipes

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
022668449 ( ALEPH )
13986162 ( OCLC )

Downloads

This item has the following downloads:


Full Text















UV REFLECTANCE OF FROSTS
COMPOSED OF WATER AND AMMONIA










By
John Gilbert Pipes
















A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY




UNIVERSITY OF FLORIDA 1972








































To my wife, Betty















ACKNOWLEDGMENTS


The author wishes to express his gratitude to Dr. R. C. Anderson, whose technical assistance, professional interest, and efforts put forth in obtaining the financial support for this study were far beyond the requirements of a committee chairman.

The author would like to thank Dr. A. E. S. Green, Dr. A. G. Smith, Dr. B. M. Leadon, Dr. M. H. Clarkson, and Dr. D. T. Williams for their efforts contributed as members of his supervisory committee.

Special acknowledgment is extended to Dr. T. G. McRae for his complete and accurate technical advice on vacuum system techniques.

The author would also like to acknowledge Mr. H. E. Stroud for his general assistance in procuring equipment and materials needed

for the construction of the experimental apparatus.

This research was supported by the National Science Foundation, Grant GA28852.















TABLE OF CONTENTS

Page

Acknowledgments iii

List of Tables v

List of Figures vi

Abstract viii

I. Introduction 1
A. Impetus 1
B. Design Considerations 3
C. Basic Results 5

II. Experimentation 7
A. Light Source 7
B. Monochromator 7
C. Frost Chamber 12
D. Photometry 13
E. Source Gases 21

III. H20 and NH Frost UV Reflectivities 23
A. host Growth Procedures 23
B. DlH3 Frost Results 25
C. H120 Frost Results 29

IV. Conclusions 90

Appendices 94
Appendix Introduction 95
Appendix 1. Light Source 96
Appendix 2. Monochromator 100
Appendix 3. Frost Chamber 102
Appendix 4. Source Gases 104
Appendix 5. Calibration of PM Tubes 108

Bibliography 112

Biographical Sketch 115









iv
















LIST OF TABLES


Table Page

1. Cross-Calibration Values 15

2. Gain-Volts Calibration Values 19

3. Reflectivity vs. Wavelength for NH3 ", 10, 11, and 12 38

4. Reflectivity vs. Wavelength for NH3 #17a and b 41

5. Reflectivity vs. Wavelength for NH3 #19a, b, and c 44

6. Reflectivity vs. Wavelength for H20 #5a and b 60

7. Reflectivity vs. Wavelength for H20 #7a and b 63

8. Reflectivity vs. Wavelength for H20 #10a and b 66

9. Reflectivity vs. Wavelength for H20 #12a and b 69

10. Reflectivity vs. Wavelength for H20 #13a, b, c, and d 74

11. Reflectivity vs. Wavelength for H20 #14a, b, c, and d 79

























v















LIST OF FIGURES


Figure Page

1. The Schematic Diagram of the Experimental
Arrangement 9

2. The Photographs of the Experimental Arrangement
and Light Source 11

3. The Cross-Calibration Curve 14

4. The Gain-Volt Calibration of PM 9553 18

5. NH3 #9: Cubic Phase 31

6. NH3 #10: Cubic Phase 33

7. NH3 #11: Cubic Phase 35

8. NH3 #12: Amorphous Phase 37

9. NH3 #17a and b: Amorphous and Cubic Phases 40

10. NH3 #19a, b, and c: Cubic Phase 43

11. NH3 Gas Absorption Coefficients 45

12. NH3 Solid Absorption Coefficients 46

13. Photographs of Cubic and Amorphous NH3 48

14. Photographs of NH3 Frosts Having Various Textures 50

15. H20 #5a and b: Cubic Phase 59

16. H20 P7a and b: Cubic Phase 62

17. H120 #10: (a), Amorphous Phase; (b), Cubic Phase 65

18. 1120 #12: (a), Amorphous Phase; (b), Cubic Phase 68

19. H120 #13A: (a), Amorphous Phase 71





vi











LIST OF FIGURES (continued) Figure Page

20. H20 #13: (b), (c), Cubic Phase; (d), NH3 Added
Over H20 Cubic 73

21. H20 #14: (a), Amorphous Phase 76

22. H20 #14: (b), (c), (d), Cubic Phase 78

23. H20 Vapor Absorption Coefficients 80

24. H20 Solid Hexagonal and Amorphous Absorption
Coefficients 81

25. BaSO4 and Stainless Steel Substrate
Reflectivities 82

26. Photographs of H20 Cubic Frosts Grown with a
Buffer Gas 84

27. Photographs of H20 Buffer-Gas-Frosts and H20
Amorphous Frost 86

28. H20 "Ball" Frost Growth Sequence 88

29. H20 Phase Change Data 89

30. Comparison of Jovian UV Albedo to NH3 Frost
Reflectivity 93

31. H2 Light Source Output as a Function of
Wavelength 99

32. Photomultiplier Photocathode Nonuniformities 110
















vii








Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy


UV REFLECTANCE OF FROSTS
COMPOSED OF WATER AND AMMONIA

BY

John Gilbert Pipes

June,, 1972

Chairman: Dr. R. C. Anderson Major Department: Aerospace Engineering

The reflectance spectra of ammonia and water frosts in the range
0 0 0
1400A to 3000A were measured near 77 K. For both gases the solid cubic and amorphous phases were examined. The cubic phase was established by slow warming of the amorphous frosts. For the ammonia frosts the cubic phase was also obtained by deposition of the gas at
0
180 K. The effects an frost reflectivity, of grain size and buffer gas during the growth period were studied. Both gases were deposited until the frosts were optically thick for 300OR radiation. The ammonia frosts have short wavelength cutoffs between 210OX and 200OX 0. Both the ammonia and water while the water frosts cut off at 1800A frosts exhibit increasing reflectivity toward shorter wavelengths. Water frosts have absorption minima centered at 2200 2075A, and
0
1925A while the ammonia frosts only show a continuum type absorption prior to the sharp cutoff. Reflectivities are less than I percent
0 0
below 1900A in the case of ammonia and below 1700A in the case of water. Annealing of the cubic phase frosts resulted in a broadening and deepening of the absorption minima.




viii















I. Introduction


A. Impetus

Middle ultraviolet spectra for most atoms and molecules in the gaseous phase, at least those relevant to the atmospheric sciences, have been recorded and in some instances data exist for the liquid and solid phase. The objective of the research report here was to determine the ultraviolet reflectivities of frosts composed of solid ammonia and water. The optical properties of atmospheric gases in the solid phase have been becoming increasingly important because of the renewed interest in the Jovian planets which has been prompted by the space program. Prime examples are the discussions by Pilcher et al. (1970), Kuiper et al. (1970a), and Kuiper et al. (1970b) of the Saturnian ring systems, believed to be covered by either an a onia or water frost. These experimenters examined the near infrared and visible region while the OAO-2 (Wallace et al., 1972) recorded the Saturnian
0
UV reflectivity down to 2250A. The temperature of the Saturnian rings is believed to be approximately 90 0 K (see Owen [1965], Harrison and Schoen [1967]).

In the case of Jupiter, Lewis (1969) has generated atmospheric models and concluded that ammonia ice clouds are present in the upper regions of the planet's atmosphere. In point of fact, it was this




-1-






-2



very prediction of solid Nil3 clouds in conjunction with the UV rocket spectrum of Jupiter obtained by Anderson et al. (1969) that provided

the impetus for this s tudy.

Anderson et al. (1969) could not explain the sharp cutoff of the
0
Jupiter albedo at 1800A using NH3 gas and attributed it to an unknown absorber. Later, employing the absorption coefficient data of Dressler and Schnepp (1960) for solid cubic ammonia, Anderson and Pipes (1971) suggested the unknown Jovian constituent to be solid cubic ammonia. Since the data of Dressler and Schnepp (1960) for solid cubic ammonia consisted of only two data points in the wavelength region of interest, it was evident more experimental work on NH3 solid was necessary.

It was thus proposed to grow NH 3 and H 20 frosts at IN 2 temperatures until they become optically thick for wavelengths near 300OX and to measure their reflectivities as far into the ultraviolet as experimentally possible. The apparatus design employed many of the experimental techniques used by the following investigators: Schnepp and Dressler (1960), studies of solid Xe, Kr, Ar; Kieffer (1968, 1969, 1970), spectral reflectance of C02-H20 frosts; and Wood et al. (1968, 1971), infrared reflectance of H20 condensed on LN2-cooled surfaces. The work of Kieffer is by far most pertinent to the understanding of frosts, since the others were examining optical properties of micron thin clear ices or at best milky ices. Nevertheless, all these publications were extremely helpful in defining the experimental

techniques employed in this study.






-3



B. Design Considerations

A number of physical properties of NH3 and H20 had to be carefully considered during the experiment design. It is well established (Seiber et al., 1970; Dressler and Schnepp, 1960; Wood et al., 1971) that water has three distinct phases as a solid. The most common is the hexagonal structure which is obtained by freezing the liquid phase or by vapor deposition above 150 0 K. A cubic structure can be formed by vapor deposition at temperatures greater than 1150 K and less than 150 0 K or by annealing the amorphous phase. Amorphous water is formed by vapor deposition at temperatures below 1150 K.

In the case of ammonia, only the cubic and amorphous forms exist. Cubic ammonia is obtained from vapor deposition above approximately 1400 K and below the melting point (195.3 0 K). Deposition at LN2 temperatures results in amorphous ammonia. The amorphous phase seems to be the least understood configuration. Black et al. (1958) and Mauer et al. (1972) have conducted x-ray diffraction experiments on amorphous ammonia. Their results indicated that ammonia has two amorphous phases and that diffraction patterns indicative of cubic ammonia sometimes appear at 40 0 K for frosts grown at LHe temperature and subsequently allowed to slowly warm. Apparently, the deposition at LN2 temperatures (77 0 K) does not assure a completely amorphous phase. This possibility was recognized and is discussed later after the phase change data are presented. Another very important conclusion by Mauer et al. (1971) is that once an amorphous phase is annealed into the cubic structure (warmed above 140 0 K) the amorphous phase






-4



cannot be obtained again by cooling the cubic to temperatures as low as LHe.

An important aspect of frost spectroscopy is the characteristic equilibrium vapor pressure. Since the frosts are grown and examined in an evacuated chamber (typically 10-4 to 10-6 torr), it is essential that their vapor pressures at IN 2 temperatures is so low that absorption by vapor is insignificant. No experimental vapor pressure data exist for NH3 and H20 at 77 0 K; however, calculated vapor pressures (see Appendix 4) are 10- 25 torr for H20 and 10 -12 torr for NH3. Thus, the effect of gaseous absorption is unimportant.

Still another important consideration is the method of forming the frost. Vapor deposition on a cryogenic surface is classified as substrate cooling and is quite different from one of nature's prime cooling mechanisms, i.e., radiative cooling. In the laboratory the

radiation is always a heat load on the frost instead of a heat loss; however, it is essentially impossible to cool every black or grey

body surrounding the frost to temperatures lower than the frost. The conductive heat load must also be considered and is no doubt much larger than the radiative load even when the frost chamber is evacuated to
-6
10 torr. This of course is assuming the walls of the chamber to be at room temperature. In short it would be difficult to simulate even approximately the frosts that exist in nature (e.g., Saturn's rings and the Martian polar cap); however, it is felt that valuable information can be extracted by growing frosts using substrate cooling.






-5



C. Basic Results

The reflectivities of fourteen separate cubic and amorphous water frosts were recorded from 300OR to 14002. All amorphous H 2 0 frosts were grown at LN2 temperatures (77 0 K) while the growth rate, concentration of buffer gas, and substrate roughness were varied. The amorphous frosts appeared milky and very fine-grained. In almost every case the reflectivity was approximately 20 to 30 percent lower than the cubic structure frosts.

A cubic water frost was obtained by allowing an amorphous frost to warm gradually ( _T_ 4 deg/min) until the change of phase occurred A t
at 150 0 K. The phase change was always accompanied by an exothermic reaction, a release of adsorbed noncondensible gases (the frost chamber pressure usually increased the order of 10 Hg over a background pressure of 5 U Hg), and an obvious increase in visible reflectivity.

The H20 amorphous frost's reflectivity is relatively constant from 300OR to 220OR at which point the reflectivity decreases 20 percent in the region from 220OR to 180OX prior to the absorption cutoff
0
at 1750A. Cubic water frosts exhibit an increasing reflectance from
0 0 0 0
3000A to 2300A and then three absorption features at 2200A, 2075A, and
0
1925A. The absorption cutoff is approximately the same for both water
0 0
phases (1800A to 1700A) with the reflectivity dropping below 1 percent from 170OR to 1400R. No data were taken for the hexagonal structured solid water.

A total of nineteen a onia frosts were grown; however, the results of the first four were inconclusive and only helped to establish experimental procedures. The same techniques for forming the H 2 0






-6



amorphous and cubic structures were employed for NH3, i.e., the amorphous frost was grown, reflectivities recorded, and then the frost was slowly warmed to the temperatures (150 to 180 0 K) required to

obtain a transformation to the cubic structure. The exothermic reaction for the NH 3 amorphous to cubic phase change wasn't as abrupt as it was for H20 so that only a small increase of the dewar warming rate was seen in the 130 0 K to 1500 K range. Since the vapor pressure of NH3 was increasing rapidly in this temperature range, no pressure

fluctuation could be observed at the phase change. An exothermic process at the NH3 phase change had also been observed by Black et al. (1958).

The NH3 cubic and amorphous have similar reflectivities from 300OR
0
to 2400A, i.e., an approximate increase of 30 percent toward shorter
0
wavelengths. The reflectivity decreases rapidly below 2300A for the
0 0
amorphous NH3 and becomes less than 1 percent between 1950A and 1400A. For the NH3 cubic the reflectance drops 30 percent from 2300X to 2275R, and then remains constant until the absorption cutoff at 21001. This
0 0
level region between 2275A and 2100A was not observed for NH3 frosts when the deposition rates were sufficiently high so that latent heat loads cause the NH3 cubic to be formed directly.

Unfortunately frosts composed of mixtures of NH3 and H120 could not be grown since NH3 is extremely corrosive in the presence of H20.

In the following chapters a description of the experimental

arrangement is given, and then the NH 3 and H20 ultraviolet reflectivities are presented and discussed. A detailed description of the instrumentation is given in the appendices.















II. Experimentation


The experimental arrangement is shown schematically in Figure 1 and photographically in Figure 2. The experimentation is best described if subdivided into the following categories: A) the H2 discharge light source, B) McPherson monochromator, C) frost chamber and cryosurface, D) Photometry, E) source gases. A. Light Source

The light source was a flow-through electrodeless discharge type. Hydrogen was used as a discharge gas at pressures between 500 and 1000 U Hg. Hydrogen exhibits a uniform continuum from 3000R to approximately 1650X so that little or no readjustment of the monochromator slits or photomultiplier tube gain was required. The source was placed sufficiently close to the entrance slit so that the optics of the monochromator were overfilled. The high temperature discharge gas was separated from the monochromator vacuum by a MgF2 window. The source proved to be quite flexible, contamination free, and extremely stable for long periods of time. The spectral distribution of the light source is given in Appendix 1.


B. Monochromator

A 0.3m scanning McPherson monochromator was employed. The

monochromatic energy requirements (never greater than lO10watts/cm 2-x)


-7-
















10
93



0


0
"4
Aj
to 0 00
.r4 "4 9: 4 0.
41 "4 44
44
54 ca
.0
41 co 41 0 1 >
$4 41 0 0

V m co "4
"4 0) :3 u a
Q 4) 44 "4 :3
93 u m 00 44 0 r-I
go "4 w C 0 0 -A CO CO
r-4 60 to w
*0 u 44 z r-4 >
r4 "4 4) "4 0 44 0 4) r4 cu
w 0 41 0 w c "i F3
.C u u 0 "4
4J 4J 4.* 41 0
r. a w 4) -r4 0 Aj 44
44 :3 $4 cc m 4 4J -A (A 4
0 c: x 4) 0 u 29 14 41
"4 U -Li k "4 44 44 V4 4)
Fi bo ri 0 m 1-4
cc C: 0 0 0 04 C14 93 44 44 0 0 -Li
34 M 0 Z 0 ;FPO u
60 w z
cc 14
-A -c4

z 0 P4 Ol E-4 :3
ri cd
4-1
co 4)
-r4 $4
M x
(1) 14 r. 1 41 1
.0 ca 4) ca (1) cu ci o :3
E-4 r-4 W :3 1
r. u
bO 4J r4 -.4 00 ca
co 0 r. 41 r 41 >
41 -A Go "4 k co 4) 41
-4 Ri to w 0 w r-4
0 bo 0 0. 41 V 0) V. 0
> 41 -v4 $4 -A "4 41
4 "4 0 4 41 ca
bo x q 0 0 41 ca "4
"4 bo a) ca 0 4 :j z Cd bo a) "4 "a
-A r-4 Fj 41 w r. > V od
Q. 01% to 4) "4 r4 Cd 34
bo :3 4 (1) > CL = 0 w
j 0 .0 w F3 ()0 >
-4 0 u :3 (U 0. 4) Z 4 41
0 4j 03 -Li 0 4.4 41
R .0 r 4 44
4j 0 0 co ca m
-0 w 4j C14 bo 4
0) 4.) r. cq I r. u 44
0) 41 od 4) V 0 0 r Q)
d) 44 0 0 p g3 Q) 44 44 4
r e% 0.44 ca 0 w 41 i w
0 r-I cn F3 Po Q. a ca :3
M co 0 4 0 8 bo bo w 44
e u u (1) 0 En 0 0 z Q) 44 "4
0 4 -0 -A m U 0 0 0. 0 "4 $4
w r-q 41 0 $4 V 0 0
41 0 0 0 > 41
p Q) X "4 0 4j 4j 4 -r4 "4
bc -0 a) "a "0 r4 .14 0 0
V. e-% co bc 4) Q) co u 0 0 0
oc (U 0 bo CO $4 4J r. 0) a) 44
u :3 'A a) 00 44 44 4 X w
Q) w 44 -0 -f 4 u ci :3
.0 a) to 4) 4J FX4 AJ (1) .0 .0
:3 4J 41 w 4-1 z 4 CD -A :j :3
-Li 0 0 r Q) 0 0 41 4J
fj 0 0 0. Q) Q, 4J 0 -4 X
0. E Q) 14 14 14 A 0 : 0 0 "4 "4 -A
4J W ICL 44 go 0 > w W

0-% 0-% #-, #-%















u
F



n













17









Lj












st A
z




















Figure 2. The Photographs of the Experimental Arrangement and Light Source



A. Experimental Arrangement

Components labeled are: 1) McPherson monochromator. 2) MKS

Baratron pressure transducer. 3) Frost chamber. 4) Preamplifier.


B. Light Source

Components labeled are: 1) H2 discharge. 2) Microware cavity.

3) Leak valve for inlet gas. 4) Outlet pipe to mechanical

vacuum pump.









-12



0
usually set the resolution at 26A; however, when the frost absorbed
0
strongly the resolution was decreased to about 50A. The scattered light from the monochromator was measured with a solar-blind photomultiplier tube and found to be negligible. To prevent the emergent beam from overfilling the photomultiplier tube used to measure the incident radiation (Io), the f/5.3 beam of the McPherson was stopped down by a pin hole located between the monochromator and the frost chamber. The resulting beam formed a 1/2" diameter spot on the Io photomultiplier tube and a 1" diameter spot on the frost.


C. Frost Chamber

A six-inch Pyrex cross formed the vacuum chamber for frost growth and photometry. The four flanged ports were used to allow monochromatic light in, to control the position of the incident light photomultiplier tube, to connect to the vacuum diffusion pump, and to support the cryosurface. The flange arrangements are shown in Figure 1. The outside of the Pyrex cross was painted black and was also covered with a doublewalled black cloth which enabled the experiments to be conducted in room lighting. Chamber pressure was monitored by a thermocouple gauge, an ionization guage, and a MKS Baratron 3mmHg transducer unit. After the frosts were grown, the chamber pressure was held at 10-6 torr while photometric data were taken.

Different cryosurfaces were used during the experimentation period. Primarily, rough and polished stainless steel frost dewars were used but for some experiments a copper dewar was substituted.






-13



The frost dewar was connected to LN2 vacuum feedthroughs to prevent O-ring freeze out and was fed from a 25 liter LN2 supply dewar. The temperature of the frost dewar was monitored by an iron-constantan thermocouple silver soldered to its front surface.


D. Photometry

Two EMR 541F-05M-18 solar-blind photomultiplier tubes (hereafter denoted PM tubes) were positioned within the frost chamber to record the incident and reflected UV radiation. These PM tubes are sensitive to radiation with wavelengths between 340O and 1400R. Each PM tube output was connected to a Fairchild solid-state preamplifier with a xl gain and a low pass filter. After amplification and filtering, the PM tube output was displayed on a chart recorder and a digital volt meter. High voltage was provided by a Fluke 0-6000 volt power supply.

One of the PM tubes could be moved remotely into the beam to record the total incoming flux while the second PM tube was mounted to collect the reflected light at approximately 100 from normal incidence (see Figure 1). The monochromator was dialed to the desired wavelength and the total incident flux was measured. The PM tube used to record this signal was then moved out of the beam and the reflected light measured. This basic procedure was continued until all wavelengths were covered. Wavelength steps of 10OR were used if the frost reflectivities were a continuum (X > 2400X) and steps of
0
25A were used if reflectivities exhibited features and absorption cutoffs.


















0
0
0
u
4)
W
0 V
0 41
r-I
OD bo
04
$4
0 W 4J
0 41
ri
0 (U 4.J
to r-I 4)
C*4 44 -0
41
bo
"4 0 -4
0 41
qt "4


-4
0
0 z C*j C*j w

w >
w
>
0
0
0 W 0
"4 4J
M

0 "4 CO
0 r-4
cc


0

0
0






bo
Oj 0 40 v C*j 0 V4
46 46 vi wi ad



2996 Nd L919 VYd






-15





Table 1

Cross -Calibration Values



Wavelength PM 6157 4 PM 9553

3000 5.83
2900 5.81
2800 5.84
2700 5.85
2600 5.78
2500 5.78
2475 5.79
2450 5.80
2425 5.81
2400 5.84
2375 5.87
2350 5.91
2325 5.96
2300 6.00
2275 6.05
2250 6.09
2225 6.14
2200 6.19
2175 6.23
2150 6.27
2125 6.31
2100 6.35
2075 6.37
2050 6.38
2025 6.39
2000 6.38
1975 6.37
1950 6.36
1925 6.34
1900 6.31
1875 6.27
1850 6.21
1825 6.16
1800 6.10
1775 6.06
1750 6.00
1725 5.96
1700 5.91







-16



Since the spectral sensitivity of the two PM tubes are differeat they had to be cross calibrated. This calibration is needed in order to calculate what the incident light would have registered on the PM tube used to measure the reflected light. The cross-calibration was determined by placing the PM tubes side by side facing the

UV beam. For some constant monochromatic incident flux the tubes were moved in and out of the beam and the resulting outputs were divided. The cross-calibration curve is shown in Figure 3 with the appropriate error bars and the values are listed in Table 1. The cross-calibration errors are essentially the total errors of the experiment and are a result of nonuniformities in the UV beam and the PM photocathodes. These nonuniformity problems were of great concern and are discussed in Appendix 5.

Calculation of the expected signal levels showed that the PM tube used to measure the incident radiation could be saturated while the PM measuring the reflected radiation would have a low signal-tonoise ratio. This problem is a direct result of the reflectance characteristics of the frost. It is assumed that the frost is a Lambert reflector and thus distributes the incident flux according to the cosine law. The angular distribution of reflected radiation from C02 cryodeposits has been measured by Smith et al. (1969) and was found to be essentially Lambertian. As positioned in the chamber, the reflected light PM tube had a collecting solid angle approximately
-3
10 times that of a hemisphere and thus the intensity of radiation on the reflected light PM tube was u 10- that of the incident light PM tube. This difficulty was overcome by operating the incident light







-17



FM tube at a reduced gain compared to the reflected light PM tube. The PM tube gain was controlled by varying the applied high voltage. The reflected light FM tube was always kept at 2950 volts while the incident light PM tube high voltage ranged from 1600 to 2800 volts.

Following this procedure of reducing the gain of one PM tube

meant that a gain-volt calibration for this particular tube had to be established so that measurements taken at a reduced gain could be accurately extrapolated to the gain at 2950 volts. This was easily accomplished by parking the PM tube in the light beam and changing the applied high voltage over a broad range (1600-2950 volts). Next, all PM tube outputs at reduced voltages were divided into the output at 2950 volts. This function is plotted in Figure 4 and tabulated in Table 2. It was found that signals at 2950 volts could be readily predicted within an accuracy of 1 percent by recording the signals at reduced voltages. In most instances the voltage of the incident light PM tube ranged between 2400 and 2800 volts, and only in cases where large incident light levels were required to obtain a respectable reflected light signal (i.e., in wavelength regions of strong frost absorption) did the high voltage have to be reduced to 1600 to 1800 volts. For this case the error is somewhat worse ( 3 percent). The 1 percent accuracy in the 2400 to 2800 volt range is attributed to the highly commendable performance of the EMR solar-blind PM tube. This gain-volt calibration was checked from time to time during the experimental period and was always found to display this remarkable accuracy.



















60

80


0
> 40





Z 2 0 7x.
U)

I- 10
0
> 8
0
S6


I- 4

z










2000 2200 2400 2600 2800


HIGH VOLTAGE

Figure 4. The Gain-Volt Calibration of PM 9553.






-19




Table 2

Gain-Volts Calibration Values

X-High Voltage Ratio Gain @ 2950 volts
Gain @ x-volts

1700 144.44

1800 92.86

1900 57.52

2000 35.14

2100 22.81

2200 14.77

2300 10.00

2400 6.67

2500 4.64

2600 3.17

2700 2.28

2800 1.60

2900 1.16

2950 1.00






-20



Since the reflected light PM tube was stationary at 10 0 from normal incidence, the total hemispheric i reflectance could only be measured by replacing the unknown reflector (NH3 and H20 frosts) by a
0
diffusely reflecting standard. This was only done for 3000A radiation to establish a hemispherical reflectance and then all the remaining wavelengths were adjusted from a relative to an absolute reflectivity.

Initially magnesium oxide was chosen as a standard. Magnesium ribbon was burned and the oxide smoke collected on aluminum or stainless steel plates. However, after some laboratory use and a review of the literature, it was clear that magnesium oxide has a number of undesirable characteristics such as rapid aging, large thicknesses are required for opacity, the powder is quite fragile, and an uncertainty in the value of total reflectivity at 3000A (see Benford et al., 1948a and 1948b). It was thus decided to replace the magnesium oxide with a barium sulfate standard. In contrast to magnesium oxide, barium sulfate has the desirable properties of small changes in reflectance with age, it can be purchased commercially as a powder or as a paint from Eastman Kodak, and the paint is fairly durable. The reflectivities of aged BaSO4 paint, fresh BaSO4 paint, and BaSO4 powder that was measured in this study are shown in Figure 25. The photometry of BaSO4 from 2000A to 8000A is discussed in detail by Billmeyer (1969), Grum and Luckey (1968), and in Kodak publications No. JJ-31 and No. JJ-32.

A summary of the photometric procedure is as follows (the incident and reflected light PM tubes are denoted by their respective serial numbers, 9553 and 6157):






-21



1) Measure I0 (X) with FM 9553 at reduced
gain (x-volts).

2) Calculate the I0 (X) which FM 9553 would
register at 2950 highvolts using the
volts-gain calibration curve.

3) Employ the cross calibration curve to
determine I0 (X) which FM 6157 would
have registered.

4) Measure Ir (W,., the reflected light
signal, with PM 6157, and divided it
by I0 (X) of step 3.

5) Adjust the relative reflectivity
(Ir [A]l/Io [Pj, to a hemispherical reflectivity by comparison to BaSO4
at 3000R.


E. Source Gases

Ultrahigh pure ammonia (99.999 percent pure) was purchased

commercially from Air Products and Chemicals Inc. and proved to be sufficiently pure. The primary foreign gas in the UHF ammonia is nitrogen which is noncondensible at 77 0 K and was thus pumped out by the diffusion pump during the frost growth. The NH3 source bottle was connected to a ballast chamber and subsequently bled into the frost chamber through a needle valve.

Obtaining pure water vapor was somewhat more difficult than NH3. A commercial still was used to produce "conductivity water" which was collected in a glass vacuum trap. This trap was cleaned with chromic acid and leached with water from the still (approximately 2000 F). After sufficient leaching, the water was collected in the trap and tested with a conductivity meter. A conductivity of 0.5 x 106 ohm- cm -1was set as an acceptable purity.







-22



The most difficult impurity to remove from water is ammonia

absorbed in the form of NH4OH. If the concentration of NH40H is very small it is totally dissociated into NH4 + and OH_ ions which are the ions measured with the conductivity meter. From order-of-magnitude calculations it can be shown (see Appendix 4) that for a conductivity of 0.5 x 10-6om cm- the NH4 +ion concentration is approximately

0.2 ppm. This level of ammonia contamination in the water is much too small to be seen in the 1120 frost spectrum. It is possible that a greater amount of NH3 was introduced into the 1120 frost by outgassing of previously adsorbed NH3. (from the chamber walls). This NH3 out-gassing could only have been the order of 10 ppm which would still be undetectable in H120 frost reflectivities.

Once the conductivity water was collected, the trap was immediately connected to the frost chamber inlet line and also to a mechanical vacuum pump line. This vacuum pump was activated so that the hot water boiled under vacuum. This process removed most of the nitrogen from the water and once the trap was valved off it contained pure water and a small amount of N2. During the 1120 frost growth the trap became cold due to the latent heat of evaporation and had to be warmed slightly in order to maintain the room temperature 1120 vapor pressure of 20 mmflg, thus insuring a uniform flow rate of 1120 vapor into the frost chamber.















III. H20 and NH3 Frost UV Reflectivities


A. Frost Growth Procedures

The techniques of controlling the growth environment for both

NH3 and H20 were strictly a result of trial and error. The understanding of how to control effectively the closely coupled parameters of latent heat, heat transfer characteristics of the frosts, vapor pressure-temperature relationships, and phases of the solids was soon found to be more difficult than the recording of photometric data. It became increasingly obvious that the initial growing conditions (flow rate, chamber pressure, substrate roughness) dictated to a large extent the growth patterns for the remaining growth period. It was also recognized that it would be difficult to define these initial conditions. Particularly for NH3, the establishment of whether the frost was in an amorphous or a cubic phase, or a combination of these two, was a major experimental problem.

From the works of Dressler and Schnepp (1960), Mauer et al. (1972), and Black et al. (1958), the techniques for obtaining an

essentially complete amorphous phase are well established. The data of Dressler and Schnepp show that the solid cubic ammonia begins to
0
absorb about 200A deeper into the UV than the amorphous solid (see Figure 12). Early data taken herein always showed that both the amorphous and cubic frosts absorb strongly between 220OR and 2000R.



-23-






-24



The early procedure for obtaining the NH3 cubic structure was to keep the substrate temperature between 1500 K and 1900 K and to have a large growth rate. The reflectivities of these cubic NH3 frosts always agreed with the amorphous frost reflectivity and thus some question of the phase was evident. This problem is discussed later after the NH3 data are presented.

Once the H20 frosts were grown, it was clear that the most convincing technique of assuring a particular phase was to grow the amorphous form first at a very slow rate and, after taking reflectance measurements, to anneal the amorphous form to a cubic form. The H20 phase change was visually obvious but temperature and pressure data were recorded to substantiate the change (see Figure 29).

Another frost growing technique established through experimentation was whether the chamber should be closed off or whether it should be pumped on with the six-inch diffusion pump during the growth period. For the NH3 and H20 sources, foreign gases were of sufficient quantity that after a three-hour growth period a sealed frost chamber would have a buffer gas present which would considerably alter the growth conditions. The basic effect of a buffer (or noncondensible) gas during frost growth is to favor the growth of any frost particles protruding from the surface since these particles see a larger concentration to the condensible gas. The obvious decision was to leave the vacuum pump open to the chamber if the effects of a buffer gas were not desired.

For most of the latest NH3 and H 2 0 frosts, the following growth procedure was followed:






-25



1) Frost chamber pumped to 106 torr (on
occasion the flanges were baked out at
T -- 1000 F).

2) Frost dewar cooled down to LN2 temperature.

3) Flow started with vacuum pump on chamber.
The chamber pressure was never above
10-3 torr during the growth period.

4) Stop flow and measure reflectance at
3000R. If it was comparable to the
BaSO4 reference, the frost was assumed
to be optically thick; if not, the flow
was turned back on.


The mass flow rate through the needle valve was never measured since it was impossible to determine what percentage of NH3 or H20 was being frozen on the cryosurface or pumped out by the diffusion pump. Also, the area of the cryosurface was poorly defined so that even if the mass flow rate was known a thickness or density measurement would have large errors. The chamber pressure was used as an indicator of the flow rate.


B. NH3 Frost Results

The UV reflectivity measurements of NH 3 cubic and amorphous

frosts are presented in Figures 5 through 10. Each figure caption gives pertinent information about the growth conditions. In addition, reflectivity data for each frost are shown in tabular form in Tables 3-5. Smooth curves were drawn through the data points listed in each table. Where zeros appear in the tables no data were taken. For comparison the gaseous NH3 data of Watanabe et al. (1953) and the solid NH3 data of Dressler and Schnepp have been reproduced in Figures 11 and 12.






-26



The most striking result was the reflectivity exceeding 100 percent for frosts optically thick at 3000R. There are two possible causes for this result; one is a consistent error in photometry due to PM calibration errors and the second is that the frosts were not Lambert reflectors for radiation in the middle ultraviolet. It must also be mentioned that the different frost thicknesses for each experiment can introduce a maximum uncertainty of 5 percent for all wavelengths. For the thicker frosts the PM tube monitoring the reflected light was closer to the frosts and thus had a larger collecting solid angle while the PM tube measuring the incident flux always collected the total light on the frost. This effect can be seen at 300OR where the reflectivity differs from frost to frost.

As for the spectral variations in reflectivity the uncertainty lies in the Lambert assumption or calibration errors. On review of the cross-calibration curve it is clear that at best the spectral variance of reflectivity could be flat or increasing toward shorter wavelengths, corresponding to negative or positive calibration errors, respectively. As discussed in Appendix 5, the errors in crosscalibration were found to stem from nonuniformities of the PM tube photocathodes and little could be done to correct this problem; however, it seems safe to conclude that the reflectivities of both NH3 and H20 cubic frosts increased toward shorter wavelengths. Supporting evidence for this conclusion is the fact that some of the amorphous frosts were indeed found to be constant in spectral reflectivity in the wavelength region of no absorption.






-27



It is just as possible that the frosts do not diffusely reflect as a Lambert surface for UV wavelengths and have a scattering phase function somewhat characteristic of a Rayleigh scattering media. This would result in a higher reflectivity near normal incidence than that of a Lambert reflector (e.g., the BaSO4 reference). Unfortunately this possibility of an unknown scattering phase function could not be examined in this experiment since it would require recording the reflectivity at all reflecting angles, a task beyond the capabilities of the system.

For the wavelength region below 240OX the ammonia frosts were easily divisible into three groups: Group 1: NH3 #9, 10, and 19 are cubic frosts that were grown directly into the cubic phase, Group 2: NH3 #11 and l7b are cubic frosts formed by warming the amorphous deposits until they were within the phase change temperature range (when data were taken the frosts were recoiled to 770 K), and Group 3: NH 3 #12 and 17a are amorphous frosts.

Group 3 shows no structure other than a continuum type cutoff from 240OR to 1950R and then was black out to 140OR (at 160OR the light source had sufficient output so that some return light could be measured and the result was a reflectivity less than 1 percent which was termed "black").

The Groups 1 and 2 were both cubic NH3 frosts but Group 2

showed a reduction of 20 percent in reflectivity between 240OR and
0
2200A and then a decrease of only 10 percent in reflectivity for the next 10OX prior to the sharp cutoff between 210OR and 2000R. Why this






-28



sharp drop in reflectivity between 2300X and 2200X and a lesser decrease between 2200X and 2100A occurred in NH3 cubic frosts formed by annealing f rom the amorphous phase could not be explained but is thought to be related to the percentage of cubic structure obtained by annealing as

opposed to growing the cubic structure directly. The physics of these phase changes clearly warrants further attention and could be best studied by x-ray diffraction techniques.

For NH3 #19, a cubic frost formed directly as a result of a

large latent heat of formation load, the reflectivity has been plotted to exemplify the repeatability of the photometric data (see Figure 10). Three complete wavelength scans of the frost were made at one-hour intervals. The vacuum pump was open to the frost chamber for the entire three-hour period and the frost was kept at 77 0 K. The uncertainty in reflectivity is approximately 1.5 percent for X > 240OR and approximately 3 percent for X < 2400X.

If the solid NH3 absorption data of Dressler and Schnepp

(Figure 12) are examined it is immediately obvious that one would anticipate a cubic frost to start abosrption about 200R deeper in the UJV than an amorphous frost. On the contrary the cubic NH3 is observed to reflect only 80R farther into UIV than the amorphous NH3 (see Figure 9) and in some cases the cutoffs are essentially identical (see Figures 6, 7, and 8).

It was because of the cubic NH3 absorption cutoff that in the early NH3 experiments the question of what phase was being examined arose. Since the results of Dressler and Schnepp give cubic NH3 absorption data only at 1875R, 1775R. and 1500R, the absorption






-29



coefficients longward of 18752 obtained by an extrapolation are uncertain. The cubic NH 3 frosts grown by annealing the amorphous frosts show higher reflectivities at shorter wavelengths than those formed directly into the cubic phase. However, the cubic NH3 of Dressler and Schnepp was formed at a high temperature, not by annealing of the amorphous NH3 and once again there appears to be conflict.

This discrepancy is difficult to analyze since amorphous NH 3 obtained by deposition at 770 K could quite possibly have contained an unknown amount of cubic NH3 and vice versa. This reasoning follows from the x-ray work of Mauer et al. (1972). At best it can only be concluded that a large percentage of what was assumed to be amorphous phase was indeed amorphous. The technique followed in this study for preparing the separate phases is given by Mauer. The only absolute assurance of phase is to examine the solid NH3 with x-ray patterns prior to measuring the UV reflectivities. Unfortunately this is beyond the scope of this research. All of these arguments are likewise applicable

to the H20 frosts.

Photographs taken during the growth of the NH3 frosts and after annealing are shown in Figures 13 and 14. In all photographs the enlargement is a factor of two. Explanations of each are given in the figure captions.


C. H120 Frost Results

The UV reflectivity measurements of H20 cubic and amorphous frosts are presented in Figures 15 through 22. The H120 amorphous and cubic results are presented separately in cases where the cubic frosts were
































0



4J
ca
0
0
"4




co



4J
z 0 4J
0
14 $4 k
0 44 z :1
-r4 4J tko bo
-4 4 r4 "4
0 0 (30 44
Z 0 0
r4 r -r4
to 0
(30 0 41
>1 co km
4 0 0
4J 14
4) 0 00 !4 0
w 0 93 ca 41
bo ''1 0 "4 4J to
-r4 $4 0 41 4J
44 :3 W 14 ca
41 $4 44
cc Q) r "4
0 vi Q) 15
$4 "4 >
bo "4 rl
0 4J Aj 00 o
14 co
ca u 44 V
29 0 0
4J CO "4
w cc $4 M 0 (v
0 bo 4J m "4 Ck.
1.4 k W 4 41
44 $4 (1) bo :3 C
4) 8 0 r-4
co 44 -Li 41 0
-rf 44 0 0 OD 0
z c 41) 1 ,
AQ

%
ca .0 tio


















A 0
..... ....



C41



40







+144
*jjjt~iji~ *Ij ~*t tI. Ut_ I l ~.~i





I Ift~ WE4 Iii I IL fJj ~ jt f ~ *~ ;j ~ iju


j. i : t
+ :~t j~tt4j ~~~4~~j1libi;If
IT-0

fli t14ij 1,a 1;: :jT~ ~:Ij
IN.-i::; ;J,0
T t L! +11. 1 0 0 0 0 0 0 0 0 0 0 0 0
z: 2w 0

( 4), ld itAI.r~























ccI


0





0



ci 4-4

w0 0
03 40


"4 0 54
01 0



0

00 0


o co


$4- 0

> 0



U 40 0d


Z -.4 0

go w



44 W

w t
0 C4 0 w
5-4 0 "4

z >

0a







-33






-0
7:. 0
0



*~F: 7: 1.l:.

0






Ci

HI- -I, Wf Iqi


cw






0



ttE 0


0~~~~~I 0I;,00 0: 0 0 0 I0 itJ 01: 0t !0 t-! .11 o

(%) tA13.3






























b 00
-4

4J0
0




00


0 c


0 w2
En 0 0L
r-. E- r a
0 0 0 <
z r 00. 0 C-4 0 1 Q) 4 0 w 0 Li 0 04 4 0 co

W- 1-4 02 "
4) 0) 0 >J 0

F34 4-J .
tv C-4 % (-4


V 04 r-44
d 4 02 J4



ca r024 ca a to 2. CU. 0 51t -A
0 r4U
r34& :2.j. I 0 4)- :5 CL 0 0

4J 0. -L
0 0i 0 4
02 4 0 0

ca 4.4 boS-








-35







... .. . . .. .. .
.4 .... ..



7.... .. 7

.. .. ..1 11TTI0

t0















~~Pi.. ....:~. ..


















!I.' li -I-








































cd

0


41
cd
41 Cd
0 Cd
.14 41 en
rn I I
"4 "a
ra a) to
C 0 0
z 0.
41) od
10 -d 9)
io -r4
0 W
r-4 V
w 44 .14
44

0
> w X
$4 > w
bo W
"4 w
44 Dt 0
0 $
C 44
4 u w
0 0 ca w
44 > r. 4
0 X
W 0 -A 4-j
m 41 41
co ci 44
X 0 ca a) 0
tko 4) 0.
0. go
$4 0 0 41L
4)
0 LW
.C 44 W 14 bo
0. 41 :1 .0 Cd 0
w .0 ::$ 41
0 to 0
0 44
Z > P4

-% -% .1%
cd 0 0) 44







-3,7









--------0

0









T400T






t11,!: hI i~t 't.







0






i Jtf r 004






-v .,



(%) ALIA~irU3i






Table 3 -38Reflectivity vs. Wavelength for NH3 #9, 10, 11, and 12



WAVELENGTH NH3 9 NH3 10 NH3 11 NH3 12


3000. 98.59 98.53 100.14 98.26
2900. 101.14 97.15 96.28 100.57
2800. 104.27 98.99 98,37 107,33
2700. 110.94 100.83 97.89 107.33
2600. 109.17 104.35 98.21 117.30
2500. 112-50 105.26 95.31 75.29
2475. 0.0 0.0 0.0 0.0
2450. 0.0 0.0 0.0 0.0
2425. 0.0 0.0 0.0 0.0
2400. 111.52 106.18 100.79 80.28
2375. 107.41 0.0 0.0 0.0
2350. 0.0 0.0 94.99 O.0
2325. 0.0 0.0 0.0 O.0
2300. 0.0 102.66 90.96 113.21
2275. 97.22 0.0 81.30 0.0
2250. 94.86 98.23 76.64 0.0
2225. 81.93 95.01 69.23 0.0
2200. 88.98 95.01 69.55 105.73
2175. 85.65 87.97 66.98 0.0
2150. 83.50 84.91 66o33 97.54
2125. 79.58 73.90 60.21 94.52
2100. 75.66 65.94 57.48 85.44
2075. 64.68 53.09 47.17 72.98
2050. 40.38 35.95 30.27 66.75
2025. 15.29 17.75 12.07 47.53
2000. 9.02 11.47 8.05 27.95
1975. 5.88 8.72 6.76 16.20
1950. 0.0 8.11 0.0 10.68





















V 4.1
0 ...4 0

00


4-4
60 0 0) 0

0 : 0

0 4.1 44
0 0) 0 e

0 01 5 4 44
0 00 50 040
Cu V. 42.4 "4 k 0 0
cc 0 1 V0
0- 0
o o oa 0
0) 0
0 0 -. 0 01
0 0 0.V
0 -A li 4

o4 0- >, -A
0i 0. 4J 0 a)0
E-4 >.- caV4
5.. "4 w
Z 0 .4 41I

co r. Cd a
> -A:
0 44
4k. 003 4
0a 01 V. 04 0 0.

r-4 0 0 0 0 0
0a 0 w. a4 0

4-1

-4 .0 (1
0 0 D 44
x C14 C. 04 0

5.04 -A.
0o 0 w 4 ~ >5...~ to 0 ~ .
Doa0 b 0 ) .41 5.0 4
"4 .44 54 -4 CN cJ
k. 0.0 w :3 Q
CO41 ) 5.4 0 44 -4 r.0c 4-4 40 .C 0 4-W
'-4 44 Q) 0 0LZ* 41 : .0 4-'J .
w .0 0o5.v4
cn 0 4J -"0 )0
$4 0 o 0 r 4. z Z w U


ca' u W







--40












.F 77:2 2





40

till 1-.



It it

C4
I a ll"








j4:f141 ~ii ]IT~~>w

H:2>





~ 0 00 0 0 0 0 00 0
iE~ N- 0 ~ ~ N






Table 4

Reflectivity vs. Wavelength for
NH3 #17a and b.



WAVELENGTH NH3 17A NH3 17B


3000. 98.60 92.96
2900. 108.07 103.58
2800. 119.52 112,88
2700. 122.51 113.21
2600. 122.18 116.03
2500. 125.66 116.37
2475. 127.82 117.36
2450. 128.32 117.69
2425. 129.31 118.86
2400. 126.33 120.68
2375. 126.49 119.35
2350. 125.33 118.86
2325. 125.83 114.87
2300. 122.67 112.38
2275. 120.85 107.57
2250. 115.10 102.42
2225. 112.55 97.94
2200. 103.25 96.28
2175. 90.30 95.12
2150. 71.55 93.46
2125. 51.96 84.83
2100. 33.53 78.02
2075. 21.75 66.90
2050. 16.77 57.93
2025. 13.61 36.52
2000. 11.12 11.95
1975. 7.47 6.47
1950. 4.48 6.31
































m cc


.0 0 to -4


4J
(d 0 >
0 r-4
cc 10 44 >,
0
4J Z Go
:3
0

>
A
0 Cd 0

0

V >
r 41 4J
ca ca m
0 cd
$4 03 :t
44 -A
0 0 -4 4J 0)
Z r-4 4) CO "4 W
44 > 13 0
k "4 C
4)
bo 4J rA
C: cd
"4 D: -A
cn 44
41
Cd V-4 :3 4J rj
-4 0 cc
Cd $4 0
0 4.J
C; "4 4) :9
4,A 14 0 r.
"4 cc 0 4 0
(A to-4 "4
0 4J 4J
to W -A
tko x m
ca 4)
!4 4J
to Co 4J
vO 4J 0
w 4)
44 .0
co 41 0
47% W -A
be 9-4 :9 -A a 4J
utim 4 0) u
41 w W 00
0 cn ca
R 0 0 $4
Go

ON
cc .0






-43






ti-A-0
... .. .... ....


... . .. ..

ip 177 -0


0~




1!IiiiIiII,... .... ....~i I










-~; 1Tt~ 't
t ~~i :'d:it U





io 1; 0 0 0 0 00
- -P m






Table 5 -44Reflectivity vs. Wavelength for
NH3 #19a, b, and c



WAVELENGTH NH3 19A NH3 198 NH3 19C


3000. 98.43 96.54 98.43
2900. 104.12 102.23 105.54
2800. 115.34 116.92 118.03
2700. 118.50 118.18 120.24
2600. 120.55 122.61 126.24
2500. 124.35 126.08 127.35
2475. 0.0 124.98 130.03
2450. 125.45 124.03 0.0
2425. 0.0 0.0 0.0
2400. 126.24 125.77 130.67
2375. 124.03 0.0 0.0
2350. 124.50 125.29 0.0
2325. 123.71 0.0 0.0
2300. 120.24 119.61 122.45
2275. 116.92 115.97 120.71
2250. 114.71 111.71 112.34
2225. 110.13 106.97 110.60
2200. 105.54 104.75 105.54
2175. 94.48 93.69 95.59
2150. 85.16 86.11 84.85
2125. 71.26 69.99 73.79
2100. 58.78 60.04 58.93
2075. 45e98 46.61 48.19
2D50. 29.23 28.12 31.76
2025. 10.90 14.22 15.96
2000. 6.48 10.74 12.48
1975. 5.85 9.95 11.85
1950. 2.05 10.43 12.48






-45





(ema 81- OTx) uol4oaS ssojo uoT4diosqV


cq
63 C-1





ell


cq

0 C9




4J

4) .0
03

00 co


0



rA 41

42 -M bb a to 0 H 4) CH


Cd

0 4J
14
0 009
OOL
oe6 009
--OOL
co V-i z


V..4

Q
14 :1










0
0 0
04

4UOT3TJJaOo UOT4djosqV

























MI







1I0









LL.

ci,~ ~ ---- -g--


--W
0









CO


2200 2000 80010N 10


WAEENT 7~







Dresie and0 Schepp 1960) 40















Figure 13. Photographs of Cubic and Amorphous NH 3


A) Amorphous Phase: Note the fine-grain size and matty
appearance.

B) Amorphous Phase: Note the two small pimples in lower
right of photo. These are believed to be cubic NH3
forming on the amorphous NH 3.

C) Cubic NH3: This cubic phase obtained by rapid deposition at 1% 1800 K. The crystals can be seen as white
specks.

D) Same as Photo C. Presented to show large grain size
compared to the amorphous.











-48-












Figure 14. Photographs of NH3 Frosts Having Various Textures

A) What was believed to be an amorphous frost with cubic growth
overlaying. Note how much brighter the pimples of cubic
appear.

B) An NH3 cubic frost after annealing at 180 0 K. Note icy
appearance.

C) NH3 cubic frost showing large grains.

D) NH3 cubic formed by annealing a cubic NH3 frost similar
to that shown in Photo C. Note how annealing greatly
reduces the voids between the grains.












-50-






-51



repeatedly annealed and photometric data recorded after each annealing, e.g., H2 #13 and 14. Each figure caption gives pertinent information to that particular frost. Reflectivity values are also listed in tabular form; where zeros appear in the tables no data were taken.

Photographs of selected H20 frosts are shown in Figures 26, 27,

and 28 with captions giving the important details. All photographs are twice the actual size. Since the frosts grown with no buffer gas were difficult to photograph most of the photographs are of buffer-gas-frosts. The latter showed substantial structure best described as a ball or cauliflowere" appearance while frosts grown without a buffer gas photographed as uniform white reflectors and little or no structure could be seen in the final prints.

For comparison, the absorption coefficients of H20 vapor (Watanabe et al., 1953) and H20 hexagonal and amorphous solids (Dressler and Schnepp, 1960) are reproduced in Figures 23 and 24.

A number of unexpected results were obtained from the H20 frosts. Most important of these are the absorption features centered at 2200K, 2075R, and 1925R. These features are more pronounced in the cubic H20 frosts but also appear in some of the amorphous frost reflectivities, e.g., H 20 #14. Since it was anticipated that both NH 3and H 20 would absorb in a continuum fashion there was immediate speculation that the three absorption features were a result of an unknown contaminant. As reviewed in Appendix 4, the technique for preparing the water was examined in detail and several alternate methods of preparation were employed. Results were always the same, i.e., the absorption features appeared in each H 20 cubic frost.






-52



Next, attention was directed to the photometry but the features in question did not appear in the reflectance measurements of BaS04 which was measured out to 180OR (see Figure 25). It was thus concluded that the photometry was correct.
0 0
The absorption at 1925A shows the greatest half width, about 100A, and is also characteristic of the manner in which the NH3 frost absorbed in the cutoff region. Since the majority of the NH3 experiments preceded the H20 experiments it was suggested that the NH3 had been adsorbed by the aluminum flanges of the frost chamber and subsequently outgassed during the H20 frost growth.

To eliminate this possibility the vacuum chamber was baked out

-6
and a 10 torr vacuum held in the chamber for several days. During the next H20 experiment (H20 #13) an amorphous frost was grown, data were taken, and then the frost was annealed above 160 0 K to obtain the cubic H20 (this was standard procedure). The reflectance measurements of the cubic H20 were recorded and the annealing procedure repeated once again. Finally a known amount of NH3 was deposited on top of the H20 frost. The result, shown in Figure 19, was an increase in re0
flectance in the 1925A absorption. The amount of NH3 admitted was approximately fifty times that calculated to be present as a contaminant.

Although this test of NH3 contamination seemed to indicate that NH3 was not causing the absorption features in the H20 frosts, the possibility still existed that the contaminant causing the features could only be influential if embedded with the lattice structure of the H20 crystals. This reasoning led directly to the theory of solid






-53



state physics and the phenomena of exciton absorption.

Before examining the possibility of explaining the H20 frost absorption features with exciton theory, several other interesting results from the H20 frost experiments are discussed.

As was observed for the NH3 frosts the reflectivity of the H20
0
cubic frosts increased from 3000A toward shorter wavelengths and in most cases exceeded 100 percent. The reasoning behind this result is identical to that given in the discussion of the NH3 frost data and thus will not be discussed further.

H20 #5 and 7 were frosts grown within a closed chamber and thus had a buffer gas present during formation. For #5 the buffer gas (N2) background pressure was about 10 11 Hg and for #7 about 100 p Hg. Both of these frosts exhibited the growth of balls ranging in sizes from < lmm to about 8mm in diameter. The effect of a buffer gas is not only to increase the growth rate of those frost grains protruding from the surface but also to increase the conductive heat load from the chamber walls. It is clear from Figure 26 that the growing balls have

small heat transfer paths to the substrate and this, coupled with the increased conductive heat load, fixed the temperature at the growth sites above the range for amorphous H20. Accordingly, both traces shown in Figures 15 and 16 are the reflectivities of cubic R 2 0 even though the substrate temperature was 77 0 K. In Figure 15 the trace labeled cubic (a) was the first scan of the "ball" frost and the cubic (b) trace was taken after annealing the frost at temperatures up to 225 0 K.






-54



Thus, the result of a buffer gas is to mix the amorphous and cubic phases and in most instances grow a cubic H20 frost overlying an amorphous H20- From Figures 15 and 16 it is also clear that since annealing of the buffer-gas-frosts increases the grain size, the shape of the absorption feature at 1925R is related to the grain size of the material. No means were available to determine an average grain size during the experiments so at best it can be concluded that an increased grain dimension also increases the strength and width of the absorption features. This grain size effect was also observed by Kieffer (1968) in the infrared region for C02 and H20 frosts.

For H20, #10, 12, 13, and 14 the six-inch diffusion pump was open to the frost chamber during the growth period so that the amorphous H20 was easily formed since all noncondensible (buffer) gases are pumped from the chamber. The amorphous H20 appeared grey and very fine-grained. It was difficult to obtain an optically thick amorphous frost since this phase of solid H20 usually grew as a translucent ice.
0
The reflectivity was checked intermittently at 3000A during the growth period and when it converged toward the BaSO4 reflectivity the growth was stopped. Typical growth periods ranged from three to five hours.

The annealing procedure outlined previously was employed to

change the amorphous phase to the cubic phase. The H20 phase change occurred at 150 0 K and was accompanied by a rise in chamber pressure (see Figure 29). A temperature rise, of the substrate, at the phase change requires an exothermic reaction within the frost. It thus follows that the energy level of the amorphous H20 must be greater, for a given temperature, than the energy level of the cubic H20 for a release of heat to occur during the phase change. This further






-55



implies that the vapor pressure of amorphous H20 should be greater than that of cubic H20 for a given temperature. When the pressure jump was observed, at the 1500 K phase change, it was felt that this could have been a result of the difference in vapor pressures for amorphous and cubic H20, Unfortunately, different vapor pressures could not be positively established for the following reasons: a) The vapor pressure for the H20 cubic has been assumed to be the same as for H20 hexagonal ice. The vapor pressure of H20 hexagonal ice has been well established (vapor pressure data taken from the Handbook of Chemistry and Physics, 44th Ed.). This assumption can be in error. b) The vapor pressure of the H20 cubic is n., 6x10_ torr at 150 0K while the outgassing of the frost chamber, when closed, increased the chamber pressure to %, 103 torr in the same time required to warm the frost from 77 0 K to 1500 K. Consequently, the H20 pressure measured with

the Baratron during the frost warm-up is an unknown partial pressure over the background pressure caused by outgassing and thus the H20 vapor pressure is only approximately determinable. c) During the growth period some noncondensible gas is always trapped within the frost. When the amorphous H20 molecules reorient and migrate during the change to an ordered structure (cubic phase) the trapped gas is released. This is probably the best explanation of the rise in chamber pressure during the phase change.

Once the cubic H20 was well established by annealing, the frost was recooled to 77 0K and during this cool-down the diffusion pump was opened to the chamber after the pressure had been reduced to n, 500 p Hg by refreezing of H 20 vapor. The refreezing was somewhat of a problem since it could cause a growth of very fine grains overlaying






-56



the annealed-large-grain frost. The pumping of vapors from the chamber during the recooling of the frost was done to minimize this problem.

For those H20 frosts grown following the above techniques the amorphous H20 reflectivities are consistently lower than the cubic H20 reflectivities (see Figures 17, 18, 19, and 20). The cubic H20 reflectivities increase 15 to 20 percent from 3000k to 2400K. This
-0.6
increase in reflectivity nicely fits a x law. As discussed by

Van de Hulst (1957) the scattering by large spherical particles is explained by a x law with a <1.0. For a = 0.6 the monodispersed particle size is found to be %, 1 0 Since the frost particles are not spherical and little is known of the change in complex index of refraction with wavelength for cubic H20, the X -06 law is only illustrative.

No consistent reason can be found to explain the reflectivity of the amorphous H20 frosts. This is expected since the opacity, grain configuration, and optical constants for the amorphous phase are poorly understood. The physical characteristics of the amorphous phase for both NH3 and H20 clearly warrants future attention.

The effects of annealing the H20 frosts were studied in H20 #14 (Figure 21). The results show a decrease in reflectivity shortward of
o 0
2500A after each annealing. The cutoff at 1800A is also influenced,
0
i.e., the annealed frosts have lower reflectivities between 1800A
0
and 1700A.
0
The possibility of attributing the absorption features at 2200A,
0 0
2075A, 1925A to a photon-exciton interaction can only be discussed in general terms since (to the knowledge of the author) experimental research, into the physics of solid water, is inconclusive at the






-57



present time. The initial suggestion that exciton absorption is causing the structure in the reflectivity of solid H20 observed in this work was made by Prinz (1972). An excellent review of exciton theory is given by Knox (1963).

The first explanation of absorption features in perfect insulating materials was given by Frankel, Peierls, and Wannier (ref. Knox, 1963) in the early 1930's. Since that time, exciton-phonon-photon interaction has been established and many elements and organic compounds studied. The most simplified exciton energy state model is the hydrogen-like model and is similar to the Rydberg series.

The features observed in the H20 frost were reviewed in light of the hydrogenic type series and no definite conclusion can be made at this time.

































Cd
W




cd
Dt
4J

04 0
z $4
44


bo 4J
z 44
to
44
44 "4


0
to

co 44
Ln cc -4
titt C> w va
0
C14 ca

x C; o
4J C14
"4 4J

u 0 "4 .0

0 Z
bo Cho 0
"4
PL4
cc co
P:
m 0 4J (1) in
m U C4
0 C13 C14
$4 w
44

4J
0 4J pq cc


cc







-59-






.. .. . -- -- .. .
0 0












''it: ~ H .it










t-. CM


I !It iio 0 0 0 00 00 ~ 0
F)~il Nil -i 0 0 t- *t )
a - -






Table 6 -60Reflectivity vs. Wavelength for
H20 #5a and b



WAVELENGTH HOH 5A HOH 58


3000. 98.79 89.89
2900. 106.62 103.77
2600. 120.86 113.03
2700. 117.48 119.97
2600. 121.93 121.43
2500. 127.98 128.69
2400. 126.74 129.05
2300. 124.07 129.41
2250. 0.0 0.0
2200. 122.11 111.61
2150. 119.08 0.0
2100. 117.48 104.31
2075. 0.0 0.0
2050. 0.0 100.57
2025. 0.0 77.43
2000. 102.53 77.25
1975. 0.0 0.0
1950. 92.56 62.83
1925. 0.0 0.0
1900. 99.50 64.97
1875. 97.19 67.64
1850. 102.71 73.34
1825. 98.26 75.12
1800. 95.41 73.34
1775. 80. 10 59.63
1750. 64.26 37.74
1725. 43.23 14.06
1700. 17.62 3.38

























44 1
V-4 (D 4) cc u r.
m 0
41 41

0 44 0 0.0 V ca
r. >1 CS AJ 4 "4 0 0) > 44
0 -r4
0 41 Q)

S-4 4) 0 0 -0
4-4 44 4) w (U DO w cd
0
"4
r-4 -r4 bO A4 m 9:
cu 0) "4 u C14 U -4
44 z 9: CO
.a to w a)
:$ w
$4 4) (d a) 44 CO
CIO 4J 44
44 4 0) CO "a C to Q) 41
10 44 0) V
r. 44 "4 W tn
co 0 :3
z
co
tko w C14
IL)
0 :L
C14 H
$4 4)


co LJ
1 (L) R
-4 0 4-J 4
Li cqo 41
"4 z


bo -,4
.14 r-4
44 0 ca
4 (10 a)
tko
W
$4 CN M co co

w 4J
4J Q) E-4 -,4
m u
0 ca 41
k 4 co .0 0
44 AJ
w 10
X $4 r. 44 .r4 41 =1 cc$ -4
0 0 Id
co

















A t .. ... .





(0







...... t p;. .:1"

I;; tjj ~ :




*t RI I LL




Ii E;. ~









.. .. ... .... ....



o 0 04 0 0 0 0 0a



COX1) AIIAIL03-1338





Table 7
-63
Reflectivity vs. Wavelength for
H 20 P7a and b,




WAVELENGTH HON 7A HOH 7B


3000. 97.01 76.18
2900. 98.43 81.22
2800. 108.58 87.04
2700. 106.27 95.76
2600. 113.39 96.30
2500. 115.34 101.46
2400. 119.97 100.21
2300. 114.45 100.93
2250. 0.0 94.34
2200. 112.67 83.66
2150. 106.27 77.25
2100. 104.66 74.23
2075. 0.0 73.69
2050o 100.21 71.02
2025. 102.17 68.71
2000. 93.63 58.56
1975. 87.93 47o53
1950. 74.76 39o87
1925. 75.83 38.09
1900. 68.71 37.38
1875. 68.17 42.01
1850. 67.64 46.99
1825. 63.55 40.76
1800. 54.29 45.57
1775. 36.67 36.49
1750. 19.22 24.92
1725. 10.15 13.35
1700. 4.63 5.52
























AID<
in
04
0%


-.4
w C
0 4J
cd


cd 0
,44 0 CM
-A
4j
ockc
00
0
93 C4
4) C14
w
co .0 41 0 co
w 4) Oi
44 0
C14
w C4
:3
41 A
0 0 co
04 0. r. 4)
54 44
rn 0
w co Ei 0
0 :1 4 0
z 0 Ei 41 41 44
0 r. m
w 0) w w co
44 Ei 0 :3 (D
0 r4 V) 0 1-4
a 4 .0 X bo
w 4) ca r.
V-4 CL tn Cd
V ca x 0) q
14 34
4J
4 co 0

bo 4J
0 vi
0
r. $4 x
44
4-4
0
0 4J 4) go
k m .0 -A
44 0 41
> R 1-4 PC
co 0 u m 0 ca
co Go :3 $4 0
bo -A -,4 0 bo bo bo

D:
ca A 4 cc w :3 1
0 w 44 0
93 4) 44
Cd 4J :3 M
u to .0 $4
to 0) 0 0
14 34 0
E-4 9 44 u z


cc 44 bo







-65






0
7- 0








40







1'' til ;I i








0 o




Itu~ it I it'
~0








N~~~ ti 00 ~ ~ D1 ~



-I -a -. ..






Table 8z6;

Reflectivity vs. Wavelength for H 0 #10a and b



WAVELENGTH HOH JOA HOH LOB


3000. 72.20 98.17
2900. 69.91 101.23
2800. 12.01 103.52
2700. 70.86 108.49
2600. 72.01 109.82
2500. 73.53 109.63
2400. 72.58 114.98
2300. 72.01 115.36
2250. 73.34 106.39
2200. 66.09 99.32
2150. 68.57 97.60
2100. 65.51 91.49
2075. 70.10 77.55
2050. 63.98 78.31
2025. 62.27 82.51
2000. 63.22 74.30
1975. 56.73 64.37
1950. 50.42 62.65
1925. 48.70 61.31
1900. 49.47 59.78
1875. 47.94 63.22
1850. 46.60 67.61
1825. 46.22 68.57
1800. 41.06 70.48
1775. 32.28 65.51
1750. 18.14 56.34
1725. 10.70 40.87
1700. 4.77 17.57






















4J
41
"4 4J
Q)
Cd ca
4.) u





Q) Cd
04 rz
4J












ca
.C -4 cc
C14 2*
0
:3 C14
0 w C-j =
.Z E-4 a)
m 0 r a
w z
0 co

as 4)
$4
co 10 z a) -Li
3: Cd
0 0) W
r4 44 (a
41 -4 Q) a)
0 0 $4 p
C4 tko 44 0 (3)
r. "4
41
0. 0 cc
w 4) D,
0 4
00 10 41 p
4) Cd Cd :$
p W -Li
0 Q) 44 Cj

u 0 41
$4 0 .14 m
::j 41 w 4J
bo X 0. u
44 0 > w "4
F&4 3.4 w 0
bo (U
a)
4-J .0 ca
w 0
0 Q)
11
f. 1 "0 0


cc u















... .. -. . ----4 0







'4 i *... :1 TO 1.:::'17 r'_ :12 li:H






w liil:%






j ~it XTA


















- --- --







Table 9

Wavelength vs. Reflectivity for
H20 #12a and b


WAVELENGTH HOH 12A HOH 12B


3000. 84.45 97.52
2900. 91.16 102.86
2800. 84.62 104.75
2700. 80.67 106.30
2600. 80.15 111.97
2500. 80.32 114.21
2400. 82.56 117.48
2300. 84.62 113.69
2250. 82.90 112.14
2200. 78.95 105.61
2150. 81.01 108.70
2100. 81.18 105.61
2075. 79.46 102.68
2050. 80.15 104.58
2025. 77.92 96.84
2000. 75.51 96.84
1975. 66.05 89.78
1950. 60.37 84.28
1925. 54.70 78.9
1900. 52.63 85.31
1875. 50.22 85.83
1850. 48.16 88.41
1825. 44.03 86.52
1800. 37.15 83.76
1775. 26.66 68.97
1750. 16.86 48.16
1725. 8.43 23.56
1700. 4.13 7.57






























44


41
m go
0 Dt



$4
03 00.4
w &n




0
4 -r4 4J
Cd PN
rA
4) 0
r-4
w 0 44
41 14
z 41
0 u ul Cd
0 W -r4 4) w
C14 .0 .0 14
cd :3 P.
u 0
4) 0 r-I
0 0 $4 .0 44
-ri 04 00
r-4 I = V
w "q 0 w
44 r-I 14
14 4) 0 w :3
:3 0
0 0. 41
41 0 m
41 0 p Lf)
14 0 >
44 4J
0 0 cc 41
$4 w 4) vs
bo V :3 44 0
4) 0 -H
41 ro .0 0 9 $4
w ra 0. 0 0 41
0 co V4 $4 06
$4 Ai bo
44 to Q. 4
$4 41 41
0 cc Dt
V4 cn 4) rA 0 0
.0 4 .0 $4 4
E-4 E-4 CO 44 0


Cd










L at





0 0





Ti1 W i T 04 : ~ IIi J


t4K: 74 4 14444



0



if ::i i if


0 o


fil I i

o~~~~~~ 0 +) 1: t- 0ti 0 Z
-: W w



















IV cn
u a V
Cd 0 w
W 93
-H
Cd $4
41
0 &J >0
-r4 V

10 0
9: w 10 0
0 0. r-4 44
C-i U 0. :3
4) 03 0 Go
w u 4)
44 V-4
co 0 ci

4) 0 $4
rA -H 41 Cd
Cd 41 41 0.
4) cd :j
V. vo cn
cd
en
w u Ln Cd a
ca r. :1 -T a)
2t 0 u to r.
41 co -H
0 :3 t Cd
u cn 0 $4
M :z .0 CW)bo
:;r Cd I
4) 44 R 4)
ci tv 0 U) 9:
0 co to -rl
to $4 4 m :t 0) 44
4 C6 E-4 4J 0 4
04 -ri 41 e- t a)
0
.r4 z 0 u 0
04 Vq 10 -ri
V >
cd r4
41 bo
-H V 0-% u ;:
A 0 ca .0 (U -H
:j C4 r-I $4
w 44 Q)
co 4) "
: : to $4 4J
$4 0 0 Cd r4
ca M 164 u Cd
>, R 44 V4 W $4
AJ to
u -H 44 4) 9:
> 0 0
-ri cc
H '0 E-4 to a) 4) A
r. 93 k 0 :3
Cd 4) V4 u cc
H r4 r. 4)
C14 44 $4 cc m 0
w w 4) u C14
$4

cc m H co 41 4)
4) vi 0 (a
04 u w 4 Cd 4
o-, V-4 u ca
"a 0 to 0 0
E-4 is w Z .0 u


cc






-73







7F-2 0



F 1-F






I T. ,


40*



~ ~ ~ ~ ~ ~ .... C4~;III7:~ r;
H;: 7717
'1 tj
7. .


viv1 ::: W. 7 .1 ..j 0... ..
+ ,
11- CM tz
M-0


.. I.....0
:2: 2;;
0

C1J


0~~~~~7 0t :000. 000

I ... .........

(%)T A.IX.LrU






-74

Table 10

Reflectivity vs. Wavelength for S2 0 #13 a, b, c, and d

HOH FROST 13


WAVELENGTH A B C D

3000. 80.49 97.86 99.87 100.87
2900. 83.00 99.87 94.02 101.03
2800. 79.83 100.87 102.54 104.21
2700. 76.49 99.53 0.0 102.04
2600. 76.32 105.21 101.37 106.88
2500. 79.66 100.87 0.0 109.55
2400. 85.67 109.89 108.88 111.22
2300. 89.01 107.88 0.0 110.22
2250. 91.35 107.21 0.0 107.55
2200. 89.34 100.87 99.20 102.04
2150. 88.84 100.87 98.20 101.87
2100. 86.00 100.20 92.18 98.53
2375. 88.18 96.36 91.85 96.36
2050. 84.33 98.20 86.67 90.85
2025. 87.01 90.85 87.51 86.51
2000. 79.49 85.34 74.31 79.83
1975. 75.98 76.32 66.80 70.47
1950. 68.64 73.65 60.45 69.47
1925. 65.46 71.31 60.79 66.47
1900. 64.63 74.31 61.12 68.47
1875. 63.96 74.65 64.13 69.47
1850. 62.12 78.66 70.64 74,15
1825. 62.12 77.32 74.31 75.98
1800o. 53,94 76.65 71.31 74.65
1775. 38.91 59.79 59,79 59.45
1750. 19.71 36.07 34.07 36.74
1725. 9.52 14.03 11.02 13.69
1700. 4.68 3.67 3.67 3.51

























W I



41 00

0 0



0 r



0
(N U)
= U) C0 Q0 0 C0
Ui) IT

0~ 41-' (
co 00
(1) 00 00


Q) 0

4-44



0 CL 0,

0 0
U) 0

C14 ) 040 a

44- 00 0N 00 0 $4 = 0) = 0u
'-4 Q) W 0 -H A
-4 E


rxw 04 -H
E- .J) -4

ca







-76

















0




tN it





i. c







w. H
Ft : : :7 ..I 0f 0ut 0t 0 0 L 0it 0 0 0 ~~~~~~fl I D Cg
-l-l(%) .LIA.LO3d30























0
4 occ
41 CD
C)
0

vo
4) 93
4) V4
0
44
93
Cd cd
14

00 4) Cd
Cd 0
4


k 44
0 w k cc
:3 41 0
44 0 44 44
co ul 44
CA P., 0
w C14 41 41
>
E-4 41
41 W -H
41 (d r-4 >
Cd 44 "A
$4
41
0 (L)
z -H bo 44


$4 Ln Cd 0 0
X I;T w 4) -H
41 k 0
0 tj 0 4)
C14 V (L) co
4) 0 10
10
Cd 0) u 4
r-f td 4)
C4 m .0 4) 44
to co 44
Cd 29 41 k -H
u 0) 10 $4 00 4) 41
:1 9: 0.44 0)
tko Cd
Cd
w co (1) 0
0 x 0 4) CD rj)
)1 0. r r. (D 7-4
44 ca 41 Ln ca
W
41 10
10 :1 co 0 0 0
w z ca z


Cd u 10



















0













0




T!. 7b P N~~~~~ ;10r. .D C 4
-t!" -- -4--







-79
Table 11

Reflectivity vs. Wavelength for
H12 0 #14 a, b, c, and d

HOH FROST 14


WAVELENGTH A a c D

3000. 105.72 97.77 102.51 90.42
2900. 109.55 109.85 0.0 94.86
2800. 109.55 108.63 113.83 110.77
2700. 110.92 115.51 0.0 113.07
2600. 115.21 116.59 115.82 116.59
2500. 114.75 122.71 0.0 118.27
2400. 119.34 122.55 120.56 118.73
2300. 119.49 120.41 117.81 115.36
2250. 118.12 121.79 114.14 108.17
2200. 110.62 111.84 112.61 103.73
2150. 112.45 116.74 112.00 104.04
2100. 111.69 115.97 108.48 99.14
2375. 106.03 115.36 104.65 99.91
2050. 104.35 109.39 105.42 96.85
2025. 108.78 106.18 96.08 93.64
2000. 104.50 101.90 94.09 86.14
1975. 92.72 95.47 85.37 82.16
1950. 88.43 93,94 85.68 74.20
1925. 85.22 89.05 80.94 74.51
1900. 84.00 92.72 82.47 72.67
1875. 84.00 89.35 84.15 76.35
1850. 82.31 85.99 84.46 79.10
1825. 84.15 87.36 83.69 79.25
1800. 77.42 83.08 76.81 70.84
1775. 65.18 70.07 58.45 49,57
1750. 43.30 46.66 29.99 23.87
1725. 23.26 20.96 8.11 7.04
1700. 8.41 6.27 3.06 2.75






-80





mo,,_Olx) uoT4oas ssojo uoT4djosqV


N N r-4


cq






v-4



co 00 0






rl



C3
CD
4-) Cd

S
0
0 to V-4

U) 4J

4-)
bb Q
"4 CH r-4

> 0 Cd r)
cl V-4 0

4J
a
0 N w
V-4

IL4
0
Cd
>
0 cli


C 4 cq




10
44




0

0 0 0 0 Oo
C> 0 0 0 0 0 0
00 t- w to w cn C14 r-4


4u9T3Tjj90o uoi4diosqV



















4'









-.7









w A_I-A







0


-zL I-I










Dresle an____p 96)








-82






.. .. ......... ....
..... ... .... .... ...



Z

A1


.. ......... :j
Ii4 4 .1
11 :
"4
I S" ;1*T ; :!1:1:7 ? i i:, I ;:: I i ;;; i;:

Ail.
7,
7
fit ci

4-i
4- !fit
.. .... ....
It 0
14 1.:
. :. :. -- .. 0
ca
PH T I! !::: +., C*j
-4"14 llld6i iW VITIr III* 'W rt, 41flil I'll+, Hi llilll '1 :11: 1: f! I ii:

3 rt 1!;;: !Ii 1 1; :111l::IV
0 U)



it itA : i
I lti ii:


-Tiff
44
Xil, I it N I i J 11, i I III! i jjrl
LN I
It R P i i H. 77::1+
T1' j" I jt;j I ilf: I I * -,;::;: :::: .

X1 t
44 4 . .
+
R ill U 1111 T' i ffi if I+
t 0:1
:--il" 14 L Ln
M4 7iH 40
Iftf ...... ::: : :. ::

o.. R if+ t"tt
+ I
14im
CIO It -111
p rT4
gli


i I LI


0 0 C-i 40 a I* W




AIIA11.03-133a












Figure 26. Photographs of H 2 0 Cubic Frosts Grown with a Bu er Gas

NOTES

A) This H 2 0 frost was grown at a slow rate with a 1000 p buffer
gas (N 2 ). The vacancy in the center righthand of the photo was once occupied by a "ball." Note the manner in which
the small balls were packed around the missing ball of frost.

B) Same photographs as "A" except for exposure time.



















-84



















































































































































































.41











Figure 27. Photographs of H 20 Buffer-Gas-Frosts
adH2 0Amorphous Frost
NOTES


A) H 0 frost grown at high deposition rate with buffer gas of
160 Hg (N ). The "ball" type frost did not develop
because of t~e nonuniform deposition obtained by a high
inlet flow of H 0 vapor. This frost is cubic even though
deposition took place at 77 0 K (see text for further
discussion).

B) Same as frost "A"; photo was taken at a different location
of dewar.

C) H 0 amorphous frost. Note the fine-grain size. Although
this frost cracked and no photometric data were taken, it depicts the optical thickness of an amorphous frost. The lighting is from the top of the picture and in the crack
which protrudes from the dewar it is clear the light
penetrates the frost for approximately 1/16 inch. The
small "ball" in this photo is the cubic H2 0 forming over
the amorphous.

D) Same frost as in photo "C" but taken earlier in the growth
period.









-86




























































































Now%













Figure 28. H 2 0 "Ball" Frost Growth Sequence


NOTES


These photographs were taken at 30-minute intervals during the growth period. The buffer gas is N 2 at 500 8. Note in photograph "A" the underlying amorphous H 2 0 formed at 77 K during start of flow.







-88




















Ar


















c D








-89








NO sci 9*9




60 0



1171 O's
0 w 0
cc
44

'A T ST L 0 $4
0 a



x 991 9 1-1 0 z
0 0 .
W
a) to
(V 4
co w 0.
-4 z
lo 5;gT *6T 41 -0
co r 0
0 4
as
4 -4 w
Xo 691 oz
'0 -0
0 0
0 ca .0 r.
04 e to
-Ao ZLI 5;*oz co p

Q) 4J
00 0
C-j p
No gL I T Z V4 4J
co co
W $4
41
0 C14 0
f3 cj m a
X UT 5'TZ
0 $4
z
4J
0 4 cn .0 C)
$40 C14 C14 w 4
x zsl a) I zz 44 4J
0 0.



98T C14 C14
0
w 54
$4 :3
0 to
w Ln r4
f7z w C*4 rT4
0 CU =L
w
P4

Xo 161 9z




NO 96T C8Z 04



00
Xo 861 COE C*4

















IV. Conclusions


The reflectivities of NH3 and H20 frosts were measured from
0 0
3000A to 1400A and the effects on reflectivity of grain growth, cubic or amorphous phase, and buffer gas were examined.

The reflectivites of the NH3 frosts are above 90 percent long0 0 0
ward of 2300A and below 1 percent from 1950A to 1400A. The absorp0 0
tion cutoff from 2200A to 1950A occurs at longer wavelengths than those expected from the absorption coefficients of solid NH3 measured by Dressler and Schnepp (1960).

The H120 frost reflectivity measurements are quite different

from anything anticipated. Based on absorption data of Dressler and Schnepp, a H120 frost should not become "black"I until approximately
0
1500A. The present study shows that if the H120 frosts are optically thick at 3000R the reflectivity will decrease to less than 10 percent
0 0
between 1800A and 1700A. This aspect is more important in relation to the study of the Jovian planets, in particular Saturn's rings.
0
Since the solar radiation at 1800A is approximately an order of magnitude
0
greater than at 1500A, detection of a solid 1120 absorption will be easier than previously thought.

It was also discovered that the H120 cubic frosts contained

absorption features which are not seen in H12 0 vapor. The possibility




-90-






-91



of these absorptions being caused by excitons was suggested.

The reflectivity of NH3 #10, a cubic frost, is plotted against the UV albedo of Jupiter (from Anderson et al., 1969) in Figure 30. The Jupiter data were normalized to the frost data at 2400g. Wavelength ranges labeled A, B, and C correspond to the first drop in the
0 0
Jovian albedo from 2200A to 2000A, to the somewhat level region from
0 0 0
2000A to 1800A and to the sharp cutoff at 1800A, respectively.

Comparison of the NH3 frost data with the Jupiter albedo in the three wavelength regions shows good agreement only in the region A. However, as stated by Anderson et al..(1969), region A can also be explained with gaseous NH3 absorption. The absorption bands of NH3 gas would not appear in the Jupiter albedo since the resolution of the rocket data was too low. In short, both gaseous and solid NH3 could explain the Jupiter albedo in the wavelength range A.

It is unfortunate that no frost data could be recorded below

1900R. Nevertheless, the frost data between 1900R and 200OR do have a trend similar to that of the Jupiter albedo.

No comment can be made concerning the wavelength range C. It should be mentioned, however, that the Jupiter albedo cutoff at
0
1800A is very similar to the H20 frost cutoff.

Finally, no comparison of the H20 frost data to the albedo of

the Saturnian ring can be made at present. The only available Saturnian UV data are that from the OAO-2 (see Wallace et al., 1972) and these data are an integrated ring plus disk albedo.




























Figure 30. Comparison of Jovian UV Albedo to NH3
Frost Reflectivity.




Full Text
-107-
It was thus concluded that the absorption features observed in
the H2O frosts were indeed due to absorption by water molecules or
as suggested by excitons.
The equilibrium vapor pressures for NH^ and 1^0 were calculated
for temperatures below those for which empirical data are available
by the following equation:
log10 P - T + b
where T is the absolute temperature, P is the pressure in mmHg and
a and b are cons tants.
This equation was fitted to the empirical data and then the
vapor pressures were calculated for temperature down to 77 K. The
final form of the vapor pressure-temperature relationships for NH^
and H^O were:
NH3: log1() P = -16.302xl02 + 9>9974
H20: log10 P = :.26-660x10 +10>5510<


MOLAR ABSORPTION COEFFICIENT (LITER MOLE CM
WAVELENGTH (&)
Figure 24. H_0 Solid- Hexagonal and Amorphous
Absorption Coefficients (from
Dressier and Schnepp, 1960).


-6-
amorphous and cubic structures were employed for NH3, i.e., the amor
phous frost was grown, reflectivities recorded, and then the frost
was slowly warmed to the temperatures (150 to 180 K) required to
obtain a transformation to the cubic structure. The exothermic re
action for the NH^ amorphous to cubic phase change wasn't as abrupt
as it was for H2O so that only a small increase of the dewar warming
rate was seen in the 130 K to 150 K range. Since the vapor pressure
of NH3 was increasing rapidly in this temperature range, no pressure
fluctuation could be observed at the phase change. An exothermic
process at the NH3 phase change had also been observed by Black et al.
(1958).
The NH3 cubic and amorphous have similar reflectivities from 3000X
o
to 2400A, i.e., an approximate increase of 30 percent toward shorter
o
wavelengths. The reflectivity decreases rapidly below 2300A for the
o o
amorphous NH3 and becomes less than 1 percent between 1950A and 1400A.
For the NH3 cubic the reflectance drops 30 percent from 2300& to 2215%.,
and then remains constant until the absorption cutoff at 2100&. This
, o o
level region between 2215k and 2100A was not observed for NH3 frosts
when the deposition rates were sufficiently high so that latent heat
loads cause the NH3 cubic to be formed directly.
Unfortunately frosts composed of mixtures of NH3 and H2O could
not be grown since NH3 is extremely corrosive in the presence of H2O.
In the following chapters a description of the experimental
arrangement is given, and then the NH^ and 1^0 ultraviolet reflectiv
ities are presented and discussed. A detailed description of the
instrumentation is given in the appendices.


-101-
The McPherson was evacuated to 10 torr with a 2" Chevron
cryo-baffle and a 2" oil diffusion pump backed by a 15 cfm Duo-Seal
forepump. The forepump and diffusion pump were separated by a Veeco
coaxial foreline trap to prevent forepump oil from backstreaming into
the monochromator. A 2" air operated gate valve separated the
McPherson from the vacuum pump, and was controlled by a safety elec
trical shutdown circuit. This combination was arranged such that in
the event of a power failure the gate valve would close and the dif
fusion pump would automatically shut off.
Scattered light from the McPherson was checked by monitoring the
output radiation with a solar-blind photomultiplier tube (PM tube)
while scanning the monochromator to wavelengths outside the range of
radiant sensitivity of the PM tube. Some signal could be seen below
1400$ if the PM tube gain was at a maximum setting and if the exit and
entrance slits of the monochromator were set to 2000 p x 2000 p (the max
imum opening of the slits). However, the slits were usually 800 p x
800 y during the frost experiments. At this setting the scattered
light for X < 1400$ was never above 5 percent of the light output for
o
1400 < X < 3400A. It is felt that if the PM tubes could have been more
solar-blina, for example the rubidium telluride photocathode long
o
wavelength cutoff is 3200A while the PM tubes used herein were cesium
o
telluride with a long wavelength cutoff at 3400A, the scattered light
component would have been essentially zero.


Figure 18. 1^0 #12: (a), Amorphous Phase; (b), Cubic Phase
NOTES
Frost growth and annealing the same as 1^0 #10.
The same procedures were followed as in H2O #10 to establish the repeatability
of observing the absorption features.
The absorption features were seen in the H2O amorphous. This may indicate that
some cubic structure was present.


-21-
1) Measure IQ (X) with PM 9553 at reduced
gain (x-volts).
2) Calculate the IG (X) which PM 9553 would
register at 2950 highvolts using the
volts-gain calibration curve.
3) Employ the cross calibration curve to
determine IQ (X) which PM 6157 would
have registered.
4) Measure Ir (X), the reflected light
signal, with PM 6157, and divided it
by Io (x) steP 3-
5) Adjust the relative reflectivity
(Ir t X]/Io [X]), to a hemispherical
reflectivity by comparison to BaS04
at 3000A.
E. Source Gases
Ultrahigh pure ammonia (99.999 percent pure) was purchased
commercially from Air Products and Chemicals Inc. and proved to be
sufficiently pure. The primary foreign gas in the UHP ammonia is
nitrogen which is noncondensible at 77 K and was thus pumped out by
the diffusion pump during the frost growth. The NH3 source bottle
was connected to a ballast chamber and subsequently bled into the
frost chamber through a needle valve.
Obtaining pure water vapor was somewhat more difficult than
NH3. A commercial still was used to produce "conductivity water"
which was collected in a glass vacuum trap. This trap was cleaned
with chromic acid and leached with water from the still (approximately
200 F). After sufficient leaching, the water was collected in the
trap and tested with a conductivity meter. A conductivity of 0.5 x
-6 -1 -1
10 ohm cm was set as an acceptable purity.


130
120
110
100
90
80
70
60
50
40
30
20
10
0


-28-
sharp drop in reflectivity between 2300& and 2200A and a lesser decrease
between 2200& and 2100A occurred in NH3 cubic frosts formed by annealing
from the amorphous phase could not be explained but is thought to be
related to the percentage of cubic structure obtained by annealing as
opposed to growing the cubic structure directly. The physics of these
phase changes clearly warrants further attention and could be best
studied by x-ray diffraction techniques.
For NH^ #19, a cubic frost formed directly as a result of a
large latent heat of formation load, the reflectivity has been plotted
to exemplify the repeatability of the photometric data (see Figure 10).
Three complete wavelength scans of the frost were made at one-hour
intervals. The vacuum pump was open to the frost chamber for the
entire three-hour period and the frost was kept at 77 K. The uncer
tainty in reflectivity is approximately 1.5 percent for \ > 2400&
and approximately 3 percent for \ < 240oX.
If the solid NH3 absorption data of Dressier and Schnepp
(Figure 12) are examined it is immediately obvious that one would
anticipate a cubic frost to start abosrption about 200? deeper in the
UV than an amorphous frost. On the contrary the cubic NH3 is observed
to reflect only 80S farther into UV than the amorphous NH3 (see Figure
9) and in some cases the cutoffs are essentially identical (see Figures
6, 7, and 8) .
It was because of the cubic NH3 absorption cutoff that in the
early NH3 experiments the question of what phase was being examined
arose. Since the results of Dressier and Schnepp give cubic NH3
absorption data only at 1875&, 1775&, and 1500&, the absorption


130
120
110
100
90
80
70
60
80
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
f
O'
I


Figure 13- Photographs of Cubic and Amorphous NH^
Amorphous Phase: Note the fine-grain size and matty
appearance.
Amorphous Phase: Note the two small pimples in lower
right of photo. These are believed to be cubic NH^
forming on the amorphous NH^.
Cubic NH3: This cubic phase obtained by rapid deposi
tion at ^ 180 K. The crystals can be seen as white
specks.
Same as Photo C. Presented to show large grain size
compared to the amorphous.


-25-
1) Frost chamber pumped to 10 torr (on
occasion the flanges were baked out at
T ~ 100 F).
2) Frost dewar cooled down to temperature.
3) Flow started with vacuum pump on chamber.
The chamber pressure was never above
10~3 torr during the growth period.
4) Stop flow and measure reflectance at
3000X. If it was comparable to the
BaSO^ reference, the frost was assumed
to be optically thick; if not, the flow
was turned back on.
The mass flow rate through the needle valve was never measured
since it was impossible to determine what percentage of NH3 or H2O
was being frozen on the cryosurface or pumped out by the diffusion
pump. Also, the area of the cryosurface was poorly defined so that
even if the mass flow rate was known a thickness or density measure
ment would have large errors. The chamber pressure was used as an
indicator of the flow rate.
B. NH^ Frost Results
The UV reflectivity measurements of NH^ cubic and amorphous
frosts are presented in Figures 5 through 10. Each figure caption gives
pertinent information about the growth conditions. In addition, re
flectivity data for each frost are shown in tabular form in Tables
3-5. Smooth curves were drawn through the data points listed in each
table. Where zeros appear in the tables no data were taken. For
comparison the gaseous NH3 data of Watanabe et al. (1953) and the
solid NH3 data of Dressier and Schnepp have been reproduced in Figures
11 and 12.


Table 4
41-
Reflectivity vs. Wavelength for
NH^ #17a and b.
WAVELENGTH
NH3 17A
NH3 17B
3000.
98.60
92. 96
2900.
108.07
103.58
2800.
119.52
112.88
2 700.
122.51
113.21
2600.
122.18
116.03
2500.
125.66
116.37
2475.
127.82
117.36
2450.
128.32
117.69
2425.
129.31
118.86
2400.
126.33
120.68
2375.
126.49
119.35
2350.
125.33
118.86
2325.
125.83
114.87
2300.
122.67
112.38
2275.
120.85
107.57
2250.
115.70
102.42
2225.
112.55
97. 94
2200.
103.25
96.28
2175.
90.30
95.12
2150.
71.55
93.46
2125.
51.96
84. 83
2100.
33.53
78.02
2075.
21.75
66.90
2050.
16.77
57.93
2025.
13.61
36.52
2000.
11.12
11.95
1975.
7.47
6.47
1950.
4.48
6.31


Absorption Coefficient (cm
CM
a
o
oo
Wavelength (X)
i
oi
i
Figure 11. NH3 Gas Absorption Coefficients (from Watanabe et al., 1953)
Absorption Cross Section (xlO


130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600
WAVELENGTH ()
28 00
3000


-51-
repeatedly annealed and photometric data recorded after each annealing,
e.g., H20 #13 and 14. Each figure caption gives pertinent information
to that particular frost. Reflectivity values are also listed in
tabular form; where zeros appear in the tables no data were taken.
Photographs of selected H20 frosts are shown in Figures 26, 27,
and 28 with captions giving the important details. All photographs are
twice the actual size. Since the frosts grown with no buffer gas were
difficult to photograph most of the photographs are of buffer-gas-frosts.
The latter showed substantial structure best described as a ball or
"cauliflower" appearance while frosts grown without a buffer gas
photographed as uniform white reflectors and little or no structure
could be seen in the final prints.
For comparison, the absorption coefficients of H20 vapor
(Watanabe et al., 1953) and H20 hexagonal and amorphous solids
(Dressier and Schnepp,I960) are reproduced in Figures 23 and 24.
A number of unexpected results were obtained from the H20 frosts.
Most important of these are the absorption features centered at 2200$,
2075$, and 1925$. These features are more pronounced in the cubic
H20 frosts but also appear in some of the amorphous frost reflectivi
ties, e.g., H2O #14. Since it was anticipated that both NH^ and 1^0
would absorb in a continuum fashion there was immediate speculation
that the three absorption features were a result of an unknown con
taminant. As reviewed in Appendix 4, the technique for preparing
the water was examined in detail and several alternate methods of pre
paration were employed. Results were always the same, i.e., the
absorption features appeared in each 1^0 cubic frost.


Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy
UV REFLECTANCE OF FROSTS
COMPOSED OF WATER AND AMMONIA
By
John Gilbert Pipes
June, 1972
Chairman: Dr. R. C. Anderson
Major Department: Aerospace Engineering
The reflectance spectra of ammonia and water frosts in the range
o o o
1400A to 3000A were measured near 77 K. For both gases the solid
cubic and amorphous phases were examined. The cubic phase was estab
lished by slow warming of the amorphous frosts. For the ammonia
frosts the cubic phase was also obtained by deposition of the gas at
o
180 K. The effects on frost reflectivity, of grain size and buffer
gas during the growth period were studied. Both gases were deposited
until the frosts were optically thick for 3000& radiation. The
ammonia frosts have short wavelength cutoffs between 2100A and 2000A
while the water frosts cut off at 1800A. Both the ammonia and water
frosts exhibit increasing reflectivity toward shorter wavelengths.
Water frosts have absorption minima centered at 2200A, 2075A, and
o
1925A while the ammonia frosts only show a continuum type absorption
prior to the sharp cutoff. Reflectivities are less than 1 percent
o o
below 1900A in the case of ammonia and below 1700A in the case of
water. Annealing of the cubic phase frosts resulted in a broaden
ing and deepening of the absorption minima.
viii


Table 6
-60-
Reflectivity vs. Wavelength for
H2O #5a and b
WAVELENGTH
HOH 5A
HOH 5B
3000.
98.79
89. 89
2900.
106.62
103.77
2600.
120.86
113.03
2700.
117.48
119.97
2600.
121.93
121.40
2500.
127.98
128.69
2400.
126.74
129.05
2300.
124.07
129.41
2250.
0.0
0.0
2200.
122.11
111.61
2150.
119.08
0.0
2100.
117.48
104.31
2C75.
0.0
0.0
2050.
0.0
100.57
2025.
0.0
77.43
2000.
102.53
77.25
1975.
0.0
0.0
1950.
92.56
62.63
1925.
0.0
0. 0
1900.
99.50
64. 97
1875.
97.19
67. 64
185C.
102.71
73.34
1825.
98.26
75. 12
1800.
95.41
73.34
1775.
80. 10
59.63
1750.
64.26
37. 74
1725.
40.23
14.06
1700.
17.62
3.38


-20-
Since the reflected light PM tube was stationary at 10 from
normal incidence, the total hemispheric'] reflectance could only be
measured by replacing the unknown reflector (NH^ and H2O frosts) by a
o
diffusely reflecting standard. This was only done for 3000A radiation
to establish a hemispherical reflectance and then all the remaining
wavelengths were adjusted from a relative to an absolute reflectivity.
Initially magnesium oxide was chosen as a standard. Magnesium
ribbon was burned and the oxide smoke collected on aluminum or stain
less steel plates. However, after some laboratory use and a review of
the literature, it was clear that magnesium oxide has a number of un
desirable characteristics such as rapid aging, large thicknesses are
required for opacity, the powder is quite fragile, and an uncertainty
o
in the value of total reflectivity at 3000A (see Benford et al., 1948a
and 1948b). It was thus decided to replace the magnesium oxide with
a barium sulfate standard. In contrast to magnesium oxide, barium
sulfate has the desirable properties of small changes in reflectance
with age, it can be purchased commercially as a powder or as a paint
from Eastman Kodak, and the paint is fairly durable. The reflectiv
ities of aged BaSO^ paint, fresh BaSO^ paint, and BaSO^ powder that
was measured in this study are shown in Figure 25, The photometry of
o o
BaSO^ from 2000A to 8000A is discussed in detail by Billmeyer (1969),
Grum and Luckey (1968), and in Kodak publications No. JJ-31 and No.
JJ-32.
A summary of the photometric procedure is as follows (the inci
dent and reflected light PM tubes are denoted by their respective serial
numbers, 9553 and 6157):


LIST OF TABLES
Table Page
1. Cross-Calibration Values 15
2. Gain-Volts Calibration Values 19
3. Reflectivity vs. Wavelength for NH3 #9, 10, 11, and 12 38
4. Reflectivity vs. Wavelength for NH3 #17a and b 41
5. Reflectivity vs. Wavelength for NH3 #19a, b, and c 44
6. Reflectivity vs. Wavelength for H2O #5a and b 60
7. Reflectivity vs. Wavelength for H2O #7a and b 63
8. Reflectivity vs. Wavelength for H2O #10a and b 66
9. Reflectivity vs. Wavelength for H2O #12a and b 69
10. Reflectivity vs. Wavelength for H2O #13a, b, c, and d 74
11. Reflectivity vs. Wavelength for H2O #14a, b, c, and d 79
v


Appendix 3
Frost Chamber
The basic layout of the 6" Pyrex cross used for a frost chamber
has been given in the main text and only a few additional points need
to be made.
The chamber outgassing rate was measured by closing the 6" gate
valve and recording the rise in chamber pressure with the MKS Baratron.
The chamber would come to about 4 y Hg in one hour starting from 10 ^
torr. This high outgassing rate was due to the necessary instrumentation
located inside. Even though all PM tube wires were Teflon insulated
and no materials other than metals were used inside the chamber, the
outgassing could not be further reduced. The chamber was leak-tested
with a He leak detector and no leaks were found. It was concluded
that outgassing was at fault. It was hoped that the outgassed con
stituents were noncondensible at 77 K and were removed by keeping the
chamber open to the vacuum pumps during the frost growth period.
An important component of the frost chamber was the aluminum
cold shield labeled part "J" in Figure 1. The shield was an aluminum
tube 5 1/2" in diameter and was attached to the frost dewar. Thermo
couple measurements at the front of the tube indicated that the tube
was cooled to ^ 100 K while closer to the dewar the temperature was even
lower. This shield served to carry a large portion of the thermal
-102-


-3-
B. Design Considerations
A number of physical properties of NH3 and H2O had to be care
fully considered during the experiment design. It is well estab
lished (Seiber et al., 1970; Dressier and Schnepp, 1960; Wood et al.,
1971) that water has three distinct phases as a solid. The most com
mon is the hexagonal structure which is obtained by freezing the liquid
phase or by vapor deposition above 150 K. A cubic structure can be
formed by vapor deposition at temperatures greater than 115 K and
less than 150 K or by annealing the amorphous phase. Amorphous water
is formed by vapor deposition at temperatures below 115 K.
In the case of ammonia, only the cubic and amorphous forms exist.
Cubic ammonia is obtained from vapor deposition above approximately
140 K and below the melting point (195.3 K). Deposition at LN2
temperatures results in amorphous ammonia. The amorphous phase seems
to be the least understood configuration. Black et al. (1958) and
Mauer et al. (1972) have conducted x-ray diffraction experiments on
amorphous ammonia. Their results indicated that ammonia has two amor
phous phases and that diffraction patterns indicative of cubic ammonia
sometimes appear at 40 K for frosts grown at LHe temperature and
subsequently allowed to slowly warm. Apparently, the deposition at
LN2 temperatures (77 K) does not assure a completely amorphous phase.
This possibility was recognized and is discussed later after the
phase change data are presented. Another very important conclusion
by Mauer et al. (1971) is that once an amorphous phase is annealed
into the cubic structure (warmed above 140 K) the amorphouse phase


-16-
Since the spectral sensitivity of the two PM tubes are differ
ent they had to be cross calibrated. This calibration is needed in
order to calculate what the incident light would have registered on
the PM tube used to measure the reflected light. The cross-calibra
tion was determined by placing the PM tubes side by side facing the
UV beam. For some constant monochromatic incident flux the tubes were
moved in and out of the beam and the resulting outputs were divided.
The cross-calibration curve is shown in Figure 3 with the appropriate
error bars and the values are listed in Table 1. The cross-calibra
tion errors are essentially the total errors of the experiment and
are a result of nonuniformities in the UV beam and the PM photocathodes.
These nonuniformity problems were of great concern and are discussed
in Appendix 5.
Calculation of the expected signal levels showed that the PM
tube used to measure the incident radiation could be saturated while
the PM measuring the reflected radiation would have a low signal-to-
noise ratio. This problem is a direct result of the reflectance
characteristics of the frost. It is assumed that the frost is a
Lambert reflector and thus distributes the incident flux according
to the cosine law. The angular distribution of reflected radiation
from CO2 cryodeposits has been measured by Smith et al. (1969) and was
found to be essentially Lambertian. As positioned in the chamber,
the reflected light PM tube had a collecting solid angle approximately
-3
10 times that of a hemisphere and thus the intensity of radiation on
_3
the reflected light PM tube was % 10 that of the incident light PM
tube. This difficulty was overcome by operating the incident light


-74-
Table 10
Reflectivity vs. Wavelength for
H20 #13 a, b, c, and d
1 HOH FROST 13
WAVELENGTH
A
B
C
D
3000.
80.49
97.86
99. 87
100.87
2900.
83.00
99.87
94.02
101.03
2800.
79.83
100.87
102.54
104.21
2700.
76.49
99.53
0.0
102.04
2600.
76.32
105.21
101.37
106.88
2500.
79.66
100.87
0.0
109.55
2400.
85.67
109.89
108.88
111.22
2300.
89.01
107.88
0.0
110.22
2250.
91.35
107.21
o.c
107.55
2200.
89.34
100.87
99.20
102.04
2150.
88.84
100.87
98.20
101.87
2100.
86.00
100.20
92.18
98. 53
2075.
86.18
96. 36
91.85
96. 36
2050.
84.33
98.20
86.67
90. 85
2025.
87.01
90.85
87.51
86.51
2000.
79.49
85.34
74.31
79. 83
1975.
75.98
76. 32
66.80
70.47
1950.
68.64
73.65
60.45
69.47
1925.
65.46
71.31
60.79
66.47
1900.
64.63
74.31
61.12
68.47
1875.
63.96
74.65
64. 13
69.47
1850.
62.12
78.66
70.64
74.15
1825.
62.12
77.32
74.31
75.98
1800.
53.94
76.65
71.31
74. 65
1775.
38.91
59.79
59.79
59.45
1750.
19.71
36.07
34.07
36. 74
1725.
9.52
14.03
11.02
13.69
1700.
4.68
3.67
3.67
3.51


2900
2800
2700
2600
2500
2475
2450
2425
2400
2375
2350
2325
2300
2275
2250
2225
2200
2175
2150
2125
2100
2075
2050
2025
2000
1975
1950
1925
1900
1875
1850
1825
1800
1775
1750
1725
-15-
Table 1
Cross-Calibration Values
PM 6157 t PM 9553
5.83
5.81
5.84
5.85
5.78
5.78
5.79
5.80
5.81
5.84
5.87
5.91
5.96
6.00
6.05
6.09
6.14
6.19
6.23
6.27
6.31
6.35
6.37
6.38
6.39
6.38
6.37
6.36
6.34
6.31
6.27
6.21
6.16
6.10
6.06
6.00
5.96
5.91


-19-
Table 2
Gain-
-Volts Calibration Values
X-High Voltage
Patio Gain @ 2950 volts
Gain @ x-volts
1700
144.44
1800
92.86
1900
57.52
2000
35.14
2100
22.81
2200
14.77
2300
10.00
2400
6.67
2500
4.64
2600
3.17
2700
2.28
2800
1.60
2900
1.16
2950
1.00


130
120
110
100
90
80
70
60
50
40
30
20
10
0
14 00
1600 1800
2 000
2200
2400 2600
WAVELENGTH (A)
2800
3 000
-33


Table 7
Reflectivity vs. Wavelength for
1^0 #7a and b
WAVELENGTH
HOH 7A
HOH 7B
3000.
97.01
76. 18
2900.
98.43
87.22
2800.
103.58
87.04
2700.
106.27
95.76
2600.
113.39
96.30
2500.
115.34
101.46
2400.
119.97
100.21
2300.
114.45
100.93
2250.
0.0
94. 34
2200.
112.67
83.66
2150.
106.27
77.25
2100.
104.66
74.23
2075.
0.0
73.69
2050.
100.21
71.02
2025.
102.17
68.71
2000.
93.63
58.56
1975.
87.93
47. 53
1950.
74.76
39.87
1925.
75.83
38.09
1900.
68.71
37.38
1875.
68.17
42.01
1650.
67.64
46. 99
1825.
63.55
40. 76
1800.
54.29
45.57
1775.
36.67
36.49
1750.
19.22
24.92
1725.
10.15
13.35
1700.
4.63
5.52


LIST OF FIGURES (continued)
Figure
Page
20.
H20 #13: (b), (c), Cubic Phase; (d), NH3 Added
Over H2O Cubic
73
21.
H2O #14: (a), Amorphous Phase
76
22.
H2O #14: (b), (c), (d), Cubic Phase
78
23.
H2O Vapor Absorption Coefficients
80
24.
H2O Solid Hexagonal and Amorphous Absorption
Coefficients
81
25.
BaSO^ and Stainless Steel Substrate
Reflectivities
82
26.
Photographs of H2O Cubic Frosts Grown with a
Buffer Gas
84
27.
Photographs of H2O Buffer-Gas-Frosts and H2O
Amorphous Frost
86
28.
H2O "Ball" Frost Growth Sequence
88
29.
H2O Phase Change Data
89
30.
Comparison of Jovian UV Albedo to NH3 Frost
Reflectivity
93
31.
H2 Light Source Output as a Function of
Wavelength
99
32.
Photomultiplier Photocathode Nonuniformities
110
vii


-22-
The most difficult impurity to remove from water is ammonia
absorbed in the form of NH^OH. If the concentration of NH^OH is very
small it is totally dissociated into NH^+ and OH ions which are the
ions measured with the conductivity meter. From order-of-magnitude
calculations it can be shown (see Appendix 4) that for a conductivity
of 0.5 x 10 6 ohm ^ cm ^ the NH^+ ion concentration is approximately
0.2 ppm. This level of ammonia contamination in the water is much too
small to be seen in the ^0 frost spectrum. It is possible that a
greater amount of NH3 was introduced into the H2O frost by out-
gassing of previously adsorbed NH^ (from the chamber walls). This
NH3 out-gassing could only have been the order of 10 ppm which would
still be undetectable in H£0 frost reflectivities.
Once the conductivity water was collected, the trap was immed
iately connected to the frost chamber inlet line and also to a mechan
ical vacuum pump line. This vacuum pump was activated so that the
hot water boiled under vacuum. This process removed most of the
nitrogen from the water and once the trap was valved off it contained
pure water and a small amount of N2. During the H2O frost growth the
trap became cold due to the latent heat of evaporation and had to be
warmed slightly in order to maintain the room temperature ^0 vapor
pressure of 20 mmHg, thus insuring a uniform flow rate of H20 vapor
into the frost chamber.


I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Mark H. Clarkson
Professor of Aerospace Engineering
This dissertation was submitted to the Dean of the College of Engi
neering and to the Graduate Council, and was accepted as partial fulfill
ment of the requirements for the degree of Doctor of Philosophy.
June, 1972
Dean, Graduate School


Table 8
-66-
Reflectivity vs. Wavelength for
1^0 #10a and b
WAVELENGTH
HOH 10A
HOH 10B
3000.
72.20
98.17
2900.
69.91
101.23
2800.
72.01
103.52
2700.
70.86
108.49
2600.
72.01
109.82
2500.
73.53
109.63
2400.
72.58
114.98
2300.
72.01
115.36
2250.
73.34
106.39
2200.
66.09
99.32
2150.
68.57
97.60
2100.
65.51
91.49
2075.
70.10
77.55
2050.
63.98
78.31
2025.
62.27
82. 51
2000.
63.22
74. 30
1975.
56.73
64.37
1950.
50.42
62.65
1925.
48.70
61.31
1900.
49.47
59. 78
1875.
47.94
63.22
1850.
46.60
67.61
1825.
46.22
68. 57
1800.
41.06
70.48
1775.
32.28
65. 51
1750.
18.14
56.34
1725.
10.70
40. 87
1700.
4.77
17.57


-52-
Next, attention was directed to the photometry but the features
in question did not appear in the reflectance measurements of BaSO^
which was measured out to 1800X (see Figure 25). It was thus con
cluded that the photometry was correct.
o o
The absorption at 1925A shows the greatest half width, about 100A,
and is also characteristic of the manner in which the NH3 frost absorbed
in the cutoff region. Since the majority of the NH-j experiments pre
ceded the H2O experiments it was suggested that the NH3 had been
adsorbed by the aluminum flanges of the frost chamber and subsequently
outgassed during the H2O frost growth.
To eliminate this possibility the vacuum chamber was baked out
-6
and a 10 torr vacuum held in the chamber for several days. During
the next H2O experiment (H2O #13) an amorphous frost was grown, data
were taken, and then the frost was annealed above 160 K to obtain the
cubic H2O (this was standard procedure). The reflectance measurements
of the cubic H2O were recorded and the annealing procedure repeated
once again. Finally a known amount of NH3 was deposited on top of the
H2O frost. The result, shown in Figure 19, was an increase in re-
o
flectance in the 1925A absorption. The amount of NH3 admitted was
approximately fifty times that calculated to be present as a contam
inant.
Although this test of NH3 contamination seemed to indicate that
NH3 was not causing the absorption features in the H2O frosts, the
possibility still existed that the contaminant causing the features
could only be influential if embedded with the lattice structure of
the H2O crystals. This reasoning led directly to the theory of solid


-91-
of these absorptions being caused by excitons was suggested.
The reflectivity of NH3 #10, a cubic frost, is plotted against
the UV albedo of Jupiter (from Anderson et al., 1969) in Figure 30.
The Jupiter data were normalized to the frost data at 2400$. Wave
length ranges labeled A, B, and C correspond to the first drop in the
o o
Jovian albedo from 2200A to 2000A, to the somewhat level region from
00 o
2000A to 1800A and to the sharp cutoff at 1800A, respectively.
Comparison of the NH3 frost data with the Jupiter albedo in the
three wavelength regions shows good agreement only in the region A.
However, as stated by Anderson et al. (1969), region A can also be
explained with gaseous NH3 absorption. The absorption bands of NH3
gas would not appear in the Jupiter albedo since the resolution of the
rocket data was too low. In short, both gaseous and solid NH3 could
explain the Jupiter albedo in the wavelength range A.
It is unfortunate that no frost data could be recorded below
1900$. Nevertheless, the frost data between 1900$ and 2000$ do have a
trend similar to that of the Jupiter albedo.
No comment can be made concerning the wavelength range C. It
should be mentioned, however, that the Jupiter albedo cutoff at
o
1800A is very similar to the H2O frost cutoff.
Finally, no comparison of the ^0 frost data to the albedo of
the Saturnian ring can be made at present. The only available Saturnian
UV data are that from the 0A0-2 (see Wallace et al., 1972) and these
data are an integrated ring plus disk albedo.


Figure 15. 1^0 #5a and b: Cubic Phase
NOTES
This frost was grown with a 10 y Hg buffer gas (N2).
Both traces are cubic H2O. Trace "b" is after frost was annealed
at T ''/225o K for about one-half hour.


TABLE OF CONTENTS
Page
Acknowledgments iii
List of Tables v
List of Figures vi
Abstract viii
I. Introduction 1
A. Impetus 1
B. Design Considerations 3
C. Basic Results 5
II. Experimentation 7
A. Light Source 7
B. Monochromator 7
C. Frost Chamber 12
D. Photometry 13
E. Source Gases 21
III. H2O and NHo Frost UV Reflectivities 23
A. Frost Growth Procedures 23
B. NH3 Frost Results 25
C. H2O Frost Results 29
IV. Conclusions 90
Appendices 94
Appendix Introduction 95
Appendix 1. Light Source 96
Appendix 2. Monochromator 100
Appendix 3. Frost Chamber 102
Appendix 4. Source Gases 104
Appendix 5. Calibration of PM Tubes 108
Bibliography H2
Biographical Sketch H5
iv


-109-
The relative sensitivity of the PM tubes was determined by
placing the tubes in the frost chamber side by side facing the UV
beam. Each tube was moved in and out of the beam and their outputs
were divided. This procedure was followed until the wavelength range
o o
1400A to 3400A was covered.
The cross-calibration was measured before the frosts were grown
and then some six months later the cross-calibration was measured
again. The second cross-calibration was in poor agreement with the
pre-cross-calibration. This discrepancy was of great concern since
all of the frost reflectivities had been calculated using the pre
cross-calibration.
It was soon recognized that only one procedure in the calibra
tion scheme had changed from the pre- to post-cross-calibration.
The PM 9553 had never been moved from its holder, since it faced the
UV beam during the calibration and frost experiments, but the PM 6157
had to be moved to the front flange of the frost chamber in order to
record the reflected radiation. During the post-cross-calibration
the PM 6157 was repositioned in the holder exactly as it was for the
pre-cross-calibration with the one exception of roll orientation.
Upon examining the effect of changing the PM 6157 in roll it
was found that a difference in signal output of up to 30 percent
resulted. It was suggested that if the UV beam and the PM photocathode
was nonuniform the result would be a change in PM tube output with a
change in roll position. To establish this fact, the UV beam was
stopped down such that a spot approximately 1/16" diameter was formed
on the face of the PM tube. The PM tube was then moved laterally so


-12-
usually set the resolution at 26A; however, when the frost absorbed
o
strongly the resolution was decreased to about 50A. The scattered
light from the monochromator was measured with a solar-blind photo
multiplier tube and found to be negligible. To prevent the emergent
beam from overfilling the photomultiplier tube used to measure the
incident radiation (I0), the f/5.3 beam of the McPherson was stopped
down by a pin hole located between the monochromator and the frost
chamber. The resulting beam formed a 1/2" diameter spot on the IQ
photomultiplier tube and a 1" diameter spot on the frost.
C. Frost Chamber
A six-inch Pyrex cross formed the vacuum chamber for frost
growth and photometry. The four flanged ports were used to allow
monochromatic light in, to control the position of the incident
light photomultiplier tube, to connect to the vacuum diffusion pump,
and to support the cryosurface. The flange arrangements are shown
in Figure 1. The outside of the Pyrex cross was painted black and
was also covered with a doublewalled black cloth which enabled the
experiments to be conducted in room lighting. Chamber pressure was
monitored by a thermocouple gauge, an ionization guage, and a MKS
Baratron 3mmHg transducer unit. After the frosts were grown, the
chamber pressure was held at 10"^ torr while photometric data were
taken.
Different cryosurfaces were used during the experimentation
period. Primarily, rough and polished stainless steel frost dewars
were used but for some experiments a copper dewar was substituted.


-84-


I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Roland C. Anderson, Chairman
Associate Professor of Aerospace
Engineering
I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy. /¡A
Alex G^ sWt
Professor of Physics and Astronomy
I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
> .
Alex E. S. Green
Graduate Research Professor in
Physics and Astronomy
I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Bernard M. Leadon
Professor of Aerospace Engineering
I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Tiic-u/ /. yy.il *-c=L i. tt J7
David T. Williams
Professor of Aerospace Engineering


Figure 7. NH^ #11: Cubic Phase
NOTES
a) The phase was established by annealing the amorphous phase at ^ 190^K.
b) Frost recooled with chamber closed to 77 K.
c) No buffer gas.
d) Amorphous was formed by very slow deposition at 77 K.
e) Growth period 3 hours 26 minutes.
-6
f) Chamber pressure % 10 torr when photometric data were recorded.
g) Note shoulder in reflectivity at 2200A.


79
Table 11
Reflectivity vs. Wavelength for
i^O #14 a, b, c, and d
1 HOH FROST 14
WAVELENGTH
A
B
C
D
3000.
105.72
97.77
102.51
90.42
2900.
109.55
109.85
0.0
94. 86
2800.
109.55
108.63
113.83
110.77
2700.
110.92
115.51
0.0
113.07
2600.
115.21
116.59
115.82
116.59
2500.
114.75
122.71
0.0
118.27
2400.
119.34
122.55
120.56
118.73
2300.
119.49
120.41
117.81
115.36
2250.
118.12
121.79
114.14
108.17
2200.
110.62
111.84
112.61
103.73
2150.
112.45
116.74
112.00
104.04
2100.
111.69
115.97
108.48
99. 14
2075.
106.03
115.36
104.65
99.91
2050.
104.35
109.39
105.42
96. 85
2025.
108.78
106.18
96.08
93.64
2000.
104.50
101.90
94.09
86. 14
1975.
92.72
95.47
85.37
82. 16
1950.
88.43
93.94
85.68
74.20
1925.
85.22
89.05
80.94
74.51
1900.
84.00
92.72
8 2.47
72.67
1875.
84.00
89. 35
84.15
76.35
1850.
82.31
85.99
84.46
79. 10
1825.
84.15
87. 36
83.69
79.25
1800.
77.42
83.08
76.81
70.84
1775.
65.18
70.07
58.45
49. 57
1750.
43.30
46.66
29.99
23. 87
1725.
23.26
20.96
8.11
7. 04
1700.
8.41
6.27
3.06
2.75


-111-
that the 1/16" diameter spot traversed its face. This was done for
both PM tubes and the results are shown in Figure 31. The distances
labeled "A" and "B" correspond to the distance the spot moved across
the face of each PM tube. If the PM photocathodes were uniform this
signal would be relatively constant as the spot traversed the face
of the tubes. As is clear from the figure this is not the case. The
fact that nonuniformities in the photocathodes of the PM tubes were
causing the errors in the cross-calibration had thus been established.
To determine the scatter in the cross-calibration values the
PM 9553 was left fixed in its holder (the same position as during the
frost experiments) while the PM 6157 was changed in roll at 45 inter
vals. The cross-calibration was measured for all eight positions of
the PM 6157. For the various positions there were some overall dif
ferences in absolute sensitivity but the important quantity is the
difference in relative "spectral" sensitivity. The eight cross
calibration curves were thus normalized at 2000X and averaged to estab
lish the final cross-calibration curve. This is shown in Figure 3
with the appropriate error bars.
All of the frost data were reduced again using the post-cross-
calibration. Prior to discovering the calibration problem the experi
mental error was approximately 2 percent; however, since no improve
ment of the UV beam and PM photocathode uniformity could be made the
maximum spectral error in the frost photometry was 5 percent.


UV REFLECTANCE OF FROSTS
COMPOSED OF WATER AND AMMONIA
By
John Gilbert Pipes
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1972

To my wife, Betty

ACKNOWLEDGMENTS
The author wishes to express his gratitude to Dr. R. C. Anderson,
whose technical assistance, professional interest, and efforts put
forth in obtaining the financial support for this study were far be
yond the requirements of a committee chairman.
The author would like to thank Dr. A. E. S. Green, Dr. A. G.
Smith, Dr. B. M. Leadon, Dr. M. H. Clarkson, and Dr. D. T. Williams
for their efforts contributed as members of his supervisory committee.
Special acknowledgment is extended to Dr. T. G. McRae for his
complete and accurate technical advice on vacuum system techniques.
The author would also like to acknowledge Mr. H. E. Stroud for
his general assistance in procuring equipment and materials needed
for the construction of the experimental apparatus.
This research was supported by the National Science Foundation,
Grant GA28852.
iii

TABLE OF CONTENTS
Page
Acknowledgments iii
List of Tables v
List of Figures vi
Abstract viii
I. Introduction 1
A. Impetus 1
B. Design Considerations 3
C. Basic Results 5
II. Experimentation 7
A. Light Source 7
B. Monochromator 7
C. Frost Chamber 12
D. Photometry 13
E. Source Gases 21
III. H2O and NHo Frost UV Reflectivities 23
A. Frost Growth Procedures 23
B. NH3 Frost Results 25
C. H2O Frost Results 29
IV. Conclusions 90
Appendices 94
Appendix Introduction 95
Appendix 1. Light Source 96
Appendix 2. Monochromator 100
Appendix 3. Frost Chamber 102
Appendix 4. Source Gases 104
Appendix 5. Calibration of PM Tubes 108
Bibliography H2
Biographical Sketch H5
iv

LIST OF TABLES
Table Page
1. Cross-Calibration Values 15
2. Gain-Volts Calibration Values 19
3. Reflectivity vs. Wavelength for NH3 #9, 10, 11, and 12 38
4. Reflectivity vs. Wavelength for NH3 #17a and b 41
5. Reflectivity vs. Wavelength for NH3 #19a, b, and c 44
6. Reflectivity vs. Wavelength for H2O #5a and b 60
7. Reflectivity vs. Wavelength for H2O #7a and b 63
8. Reflectivity vs. Wavelength for H2O #10a and b 66
9. Reflectivity vs. Wavelength for H2O #12a and b 69
10. Reflectivity vs. Wavelength for H2O #13a, b, c, and d 74
11. Reflectivity vs. Wavelength for H2O #14a, b, c, and d 79
v

LIST OF FIGURES
Figure Page
1. The Schematic Diagram of the Experimental
Arrangement 9
2. The Photographs of the Experimental Arrangement
and Light Source 11
3. The Cross-Calibration Curve 14
4. The Gain-Volt Calibration of PM 9553 18
5. NH3 #9: Cubic Phase 31
6. NH3 #10: Cubic Phase 33
7. NH3 #11: Cubic Phase 35
8. NH3 #12: Amorphous Phase 37
9. NH3 #17a and b: Amorphous and Cubic Phases 40
10. NH3 #19a, b, and c: Cubic Phase 43
11. NH3 Gas Absorption Coefficients 45
12. NH3 Solid Absorption Coefficients 46
13. Photographs of Cubic and Amorphous NH3 48
14. Photographs of NH3 Frosts Having Various Textures 50
15. H2O #5a and b: Cubic Phase 59
16. H20 #7a and b: Cubic Phase 62
17. H20 #10: (a), Amorphous Phase; (b), Cubic Phase 65
18. H20 #12: (a), Amorphous Phase; (b), Cubic Phase 68
19. H20 #13A: (a), Amorphous Phase 71
vi

LIST OF FIGURES (continued)
Figure
Page
20.
H20 #13: (b), (c), Cubic Phase; (d), NH3 Added
Over H2O Cubic
73
21.
H2O #14: (a), Amorphous Phase
76
22.
H2O #14: (b), (c), (d), Cubic Phase
78
23.
H2O Vapor Absorption Coefficients
80
24.
H2O Solid Hexagonal and Amorphous Absorption
Coefficients
81
25.
BaSO^ and Stainless Steel Substrate
Reflectivities
82
26.
Photographs of H2O Cubic Frosts Grown with a
Buffer Gas
84
27.
Photographs of H2O Buffer-Gas-Frosts and H2O
Amorphous Frost
86
28.
H2O "Ball" Frost Growth Sequence
88
29.
H2O Phase Change Data
89
30.
Comparison of Jovian UV Albedo to NH3 Frost
Reflectivity
93
31.
H2 Light Source Output as a Function of
Wavelength
99
32.
Photomultiplier Photocathode Nonuniformities
110
vii

Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy
UV REFLECTANCE OF FROSTS
COMPOSED OF WATER AND AMMONIA
By
John Gilbert Pipes
June, 1972
Chairman: Dr. R. C. Anderson
Major Department: Aerospace Engineering
The reflectance spectra of ammonia and water frosts in the range
o o o
1400A to 3000A were measured near 77 K. For both gases the solid
cubic and amorphous phases were examined. The cubic phase was estab
lished by slow warming of the amorphous frosts. For the ammonia
frosts the cubic phase was also obtained by deposition of the gas at
o
180 K. The effects on frost reflectivity, of grain size and buffer
gas during the growth period were studied. Both gases were deposited
until the frosts were optically thick for 3000& radiation. The
ammonia frosts have short wavelength cutoffs between 2100A and 2000A
while the water frosts cut off at 1800A. Both the ammonia and water
frosts exhibit increasing reflectivity toward shorter wavelengths.
Water frosts have absorption minima centered at 2200A, 2075A, and
o
1925A while the ammonia frosts only show a continuum type absorption
prior to the sharp cutoff. Reflectivities are less than 1 percent
o o
below 1900A in the case of ammonia and below 1700A in the case of
water. Annealing of the cubic phase frosts resulted in a broaden
ing and deepening of the absorption minima.
viii

I. Introduction
A. Impetus
Middle ultraviolet spectra for most atoms and molecules in the
gaseous phase, at least those relevant to the atmospheric sciences,
have been recorded and in some instances data exist for the liquid
and solid phase. The objective of the research report here was to
determine the ultraviolet reflectivities of frosts composed of solid
ammonia and water. The optical properties of atmospheric gases in the
solid phase have been becoming increasingly important because of the
renewed interest in the Jovian planets which has been prompted by the
space program. Prime examples are the discussions by Pilcher et al.
(1970), Kuiper et al. (1970a), and Kuiper et al. (1970b) of the Saturn
ian ring systems, believed to be covered by either an ammonia or water
frost. These experimenters examined the near infrared and visible
region while the 0A0-2 (Wallace et al., 1972) recorded the Saturnian
o
UV reflectivity down to 2250A. The temperature of the Saturnian rings
is believed to be approximately 90 K (see Owen [1965], Harrison and
Schoen [1967]).
In the case of Jupiter, Lewis (1969) has generated atmospheric
models and concluded that ammonia ice clouds are present in the upper
regions of the planet's atmosphere. In point of fact, it was this
-1-

-2-
very prediction of solid NH^ clouds in conjunction with the UV rocket
spectrum of Jupiter obtained by Anderson et al. (1969) that provided
the impetus for this study.
Anderson et al. (1969) could not explain the sharp cutoff of the
o
Jupiter albedo at 1800A using NH3 gas and attributed it to an unknown
absorber. Later, employing the absorption coefficient data of
Dressier and Schnepp (1960) for solid cubic ammonia, Anderson and Pipes
(1971) suggested the unknown Jovian constituent to be solid cubic
ammonia. Since the data of Dressier and Schnepp (1960) for solid cubic
ammonia consisted of only two data points in the wavelength region of
interest, it was evident more experimental work on NH3 solid was
necessary.
It was thus proposed to grow NH^ and f^O frosts at LN^ tempera
tures until they become optically thick for wavelengths near 3000^
and to measure their reflectivities as far into the ultraviolet as
experimentally possible. The apparatus design employed many of the
experimental techniques used by the following investigators: Schnepp
and Dressier (1960), studies of solid Xe, Kr, Ar; Kieffer (1968, 1969,
1970), spectral reflectance of CO2-H2O frosts; and Wood et al. (1968,
1971), infrared reflectance of condensed on L^-cooled surfaces.
The work of Kieffer is by far most pertinent to the understanding of
frosts, since the others were examining optical properties of micron
thin clear ices or at best milky ices. Nevertheless, all these pub
lications were extremely helpful in defining the experimental
techniques employed in this study.

-3-
B. Design Considerations
A number of physical properties of NH3 and H2O had to be care
fully considered during the experiment design. It is well estab
lished (Seiber et al., 1970; Dressier and Schnepp, 1960; Wood et al.,
1971) that water has three distinct phases as a solid. The most com
mon is the hexagonal structure which is obtained by freezing the liquid
phase or by vapor deposition above 150 K. A cubic structure can be
formed by vapor deposition at temperatures greater than 115 K and
less than 150 K or by annealing the amorphous phase. Amorphous water
is formed by vapor deposition at temperatures below 115 K.
In the case of ammonia, only the cubic and amorphous forms exist.
Cubic ammonia is obtained from vapor deposition above approximately
140 K and below the melting point (195.3 K). Deposition at LN2
temperatures results in amorphous ammonia. The amorphous phase seems
to be the least understood configuration. Black et al. (1958) and
Mauer et al. (1972) have conducted x-ray diffraction experiments on
amorphous ammonia. Their results indicated that ammonia has two amor
phous phases and that diffraction patterns indicative of cubic ammonia
sometimes appear at 40 K for frosts grown at LHe temperature and
subsequently allowed to slowly warm. Apparently, the deposition at
LN2 temperatures (77 K) does not assure a completely amorphous phase.
This possibility was recognized and is discussed later after the
phase change data are presented. Another very important conclusion
by Mauer et al. (1971) is that once an amorphous phase is annealed
into the cubic structure (warmed above 140 K) the amorphouse phase

-4-
cannot be obtained again by cooling the cubic to temperatures as low
as LHe.
An important aspect of frost spectroscopy is the characteristic
equilibrium vapor pressure. Since the frosts are grown and examined
-4 -6
in an evacuated chamber (typically 10 to 10 torr), it is essential
that their vapor pressures at temperatures is so low that absorp
tion by vapor is insignificant. No experimental vapor pressure data
exist for NH3 and H2O at 77 K; however, calculated vapor pressures
-25 -12
(see Appendix 4) are 10 torr for H2O and 10 torr for NH3. Thus,
the effect of gaseous absorption is unimportant.
Still another important consideration is the method of forming
the frost. Vapor deposition on a cryogenic surface is classified as
substrate cooling and is quite different from one of nature's prime
cooling mechanisms, i.e., radiative cooling. In the laboratory the
radiation is always a heat load on the frost instead of a heat loss;
however, it is essentially impossible to cool every black or grey
body surrounding the frost to temperatures lower than the frost. The
conductive heat load must also be considered and is no doubt much larger
than the radiative load even when the frost chamber is evacuated to
-6
10 torr. This of course is assuming the walls of the chamber to
be at room temperature. In short it would be difficult to simulate
even approximately the frosts that exist in nature (e.g., Saturn's rings
and the Martian polar cap); however, it is felt that valuable informa
tion can be extracted by growing frosts using substrate cooling.

-5-
C. Basic Results
The reflectivities of fourteen separate cubic and amorphous water
frosts were recorded from 3000X to 1400?. All amorphous ^0 frosts
were grown at LN2 temperatures (77 K) while the growth rate, concen
tration of buffer gas, and substrate roughness were varied. The amor
phous frosts appeared milky and very fine-grained. In almost every
case the reflectivity was approximately 20 to 30 percent lower than the
cubic structure frosts.
A cubic water frost was obtained by allowing an amorphous frost
to warm gradually T % 4 deg/min) until the change of phase occurred
A t
at 150 K. The phase change was always accompanied by an exothermic
reaction, a release of adsorbed noncondensible gases (the frost chamber
pressure usually increased the order of 10 y Hg over a background
pressure of 5 y Hg), and an obvious increase in visible reflectivity.
The H2O amorphous frost's reflectivity is relatively constant
from 3000X to 2200X at which point the reflectivity decreases 20 per
cent in the region from 2200^ to 180C)£ prior to the absorption cutoff
o
at 1750A. Cubic water frosts exhibit an increasing reflectance from
00 00
3000A to 2300A and then three absorption features at 2200A, 2075A, and
o
1925A. The absorption cutoff is approximately the same for both water
o o
phases (1800A to 1700A) with the reflectivity dropping below 1 percent
from 170oS to 140oX. No data were taken for the hexagonal structured
solid water.
A total of nineteen ammonia frosts were grown; however, the re
sults of the first four were inconclusive and only helped to establish
experimental procedures. The same techniques for forming the H20

-6-
amorphous and cubic structures were employed for NH3, i.e., the amor
phous frost was grown, reflectivities recorded, and then the frost
was slowly warmed to the temperatures (150 to 180 K) required to
obtain a transformation to the cubic structure. The exothermic re
action for the NH^ amorphous to cubic phase change wasn't as abrupt
as it was for H2O so that only a small increase of the dewar warming
rate was seen in the 130 K to 150 K range. Since the vapor pressure
of NH3 was increasing rapidly in this temperature range, no pressure
fluctuation could be observed at the phase change. An exothermic
process at the NH3 phase change had also been observed by Black et al.
(1958).
The NH3 cubic and amorphous have similar reflectivities from 3000X
o
to 2400A, i.e., an approximate increase of 30 percent toward shorter
o
wavelengths. The reflectivity decreases rapidly below 2300A for the
o o
amorphous NH3 and becomes less than 1 percent between 1950A and 1400A.
For the NH3 cubic the reflectance drops 30 percent from 2300& to 2215%.,
and then remains constant until the absorption cutoff at 2100&. This
, o o
level region between 2215k and 2100A was not observed for NH3 frosts
when the deposition rates were sufficiently high so that latent heat
loads cause the NH3 cubic to be formed directly.
Unfortunately frosts composed of mixtures of NH3 and H2O could
not be grown since NH3 is extremely corrosive in the presence of H2O.
In the following chapters a description of the experimental
arrangement is given, and then the NH^ and 1^0 ultraviolet reflectiv
ities are presented and discussed. A detailed description of the
instrumentation is given in the appendices.

II. Experimentation
The experimental arrangement is shown schematically in Figure 1
and photographically in Figure 2. The experimentation is best de
scribed if subdivided into the following categories: A) the H£ dis
charge light source, B) McPherson monochromator, C) frost chamber and
cryosurface, D) Photometry, E) source gases.
A. Light Source
The light source was a flow-through electrodeless discharge type.
Hydrogen was used as a discharge gas at pressures between 500 and
1000 y Hg. Ifydrogen exhibits a uniform continuum from 3000^ to ap
proximately 1650& so that little or no readjustment of the monochrom
ator slits or photomultiplier tube gain was required. The source was
placed sufficiently close to the entrance slit so that the optics of
the monochromator were overfilled. The high temperature discharge
gas was separated from the monochromator vacuum by a MgF£ window. The
source proved to be quite flexible, contamination free, and extremely
stable for long periods of time. The spectral distribution of the
light source is given in Appendix 1.
B. Monochromator
A 0.3m scanning McPherson monochromator was employed. The
monochromatic energy requirements (never greater than 10~10watts/cm2-#)
-7-

Figure 1. The Schematic Diagram of the
Experimental Arrangement
A) Top Flange: Components: 1) high voltage and
PM tube electrical feedthrough. 2) PM tube
remote control. 3) thermocouple and ioniza
tion gauge.
B) Front Flange: Components: 1) all gas inlet
pipes. 2) holder for PM tube monitoring re
flected radiation. 3) observation port.
A) outlet to MRS Baratron pressure transducer.
C) Back Flange: Components: 1) LN2 inlet and
outlet feedthroughs. 2) temperature thermo
couple feedthrough. 3) roughing line vacuum
valve.
D) Six inch air operated gate valve.
E) Six inch chevron cryo-baffle.
F) Six inch oil diffusion pump.
G) Cryo-surface
H) PM tube monitoring incident radiation.
I) PM tube monitoring reflected radiation.
J) Aluminum cold shiled.
K) Monochromatic incident radiation.
L) Frost and reflected radiation
M) LN2 reservoir.
N) LN2 inlet.
O) LN2 feedthroughs.
P) Connection flange containing pin hole and
MgF£ window.
Q) McPherson monochromator.
R) Two inch chevron cryo-baffle.
S) Two inch oil diffusion pump.
T) Outlet to mechanical vacuum pump
U) One-half inch aluminum base plate.


Figure 2. The Photographs of the Experimental
Arrangement and Light Source
A. Experimental Arrangement
Components labeled are: 1) McPherson monochromator. 2) MKS
Baratron pressure transducer. 3) Frost chamber. 4) Preampli
fier.
B. Light Source
Components labeled are: 1) H£ discharge. 2) Microware cavity.
3) Leak valve for inlet gas. 4) Outlet pipe to mechanical
vacuum pump.

-11-

-12-
usually set the resolution at 26A; however, when the frost absorbed
o
strongly the resolution was decreased to about 50A. The scattered
light from the monochromator was measured with a solar-blind photo
multiplier tube and found to be negligible. To prevent the emergent
beam from overfilling the photomultiplier tube used to measure the
incident radiation (I0), the f/5.3 beam of the McPherson was stopped
down by a pin hole located between the monochromator and the frost
chamber. The resulting beam formed a 1/2" diameter spot on the IQ
photomultiplier tube and a 1" diameter spot on the frost.
C. Frost Chamber
A six-inch Pyrex cross formed the vacuum chamber for frost
growth and photometry. The four flanged ports were used to allow
monochromatic light in, to control the position of the incident
light photomultiplier tube, to connect to the vacuum diffusion pump,
and to support the cryosurface. The flange arrangements are shown
in Figure 1. The outside of the Pyrex cross was painted black and
was also covered with a doublewalled black cloth which enabled the
experiments to be conducted in room lighting. Chamber pressure was
monitored by a thermocouple gauge, an ionization guage, and a MKS
Baratron 3mmHg transducer unit. After the frosts were grown, the
chamber pressure was held at 10"^ torr while photometric data were
taken.
Different cryosurfaces were used during the experimentation
period. Primarily, rough and polished stainless steel frost dewars
were used but for some experiments a copper dewar was substituted.

-13-
The frost dewar was connected to LN£ vacuum feedthroughs to prevent
O-ring freeze out and was fed from a 25 liter LN^ supply dewar. The
temperature of the frost dewar was monitored by an iron-constantan
thermocouple silver soldered to its front surface.
D. Photometry
Two EMR 541F-05M-18 solar-blind photomultiplier tubes (hereafter
denoted PM tubes) were positioned within the frost chamber to record
the incident and reflected UV radiation. These PM tubes are sensitive
to radiation with wavelengths between 340oX and 140oX. Each PM tube
output was connected to a Fairchild solid-state preamplifier with a
xl gain and a low pass filter. After amplification and filtering,
the PM tube output was displayed on a chart recorder and a digital
volt meter. High voltage was provided by a Fluke 0-6000 volt power
supply.
One of the PM tubes could be moved remotely into the beam to
record the total incoming flux while the second PM tube was mounted
to collect the reflected light at approximately 10 from normal in
cidence (see Figure 1). The monochromator was dialed to the desired
wavelength and the total incident flux was measured. The PM tube
used to record this signal was then moved out of the beam and the
reflected light measured. This basic procedure was continued until
all wavelengths were covered. Wavelength steps of 100& were used if
the frost reflectivities were a continuum (X > 2400^) and steps of
o
25A were used if reflectivities exhibited features and absorption
cutoffs.

PM 6157 -7- PM 9553
m i i i i
6.6
6.4
6.2
6.0
5.9
5.6
5.4
5.2
5.0
I I I I I I I I I
J I I I I i I I I I I I I I I I
1600 1800 2000 2200 2400 2600
WAVELENGTH ()
2800
3000
Figure 3. The Cross-Calibration Curve. FM 6157 is the reflected light detector
and PM 9553 is the incident light detector.
i
-14

2900
2800
2700
2600
2500
2475
2450
2425
2400
2375
2350
2325
2300
2275
2250
2225
2200
2175
2150
2125
2100
2075
2050
2025
2000
1975
1950
1925
1900
1875
1850
1825
1800
1775
1750
1725
-15-
Table 1
Cross-Calibration Values
PM 6157 t PM 9553
5.83
5.81
5.84
5.85
5.78
5.78
5.79
5.80
5.81
5.84
5.87
5.91
5.96
6.00
6.05
6.09
6.14
6.19
6.23
6.27
6.31
6.35
6.37
6.38
6.39
6.38
6.37
6.36
6.34
6.31
6.27
6.21
6.16
6.10
6.06
6.00
5.96
5.91

-16-
Since the spectral sensitivity of the two PM tubes are differ
ent they had to be cross calibrated. This calibration is needed in
order to calculate what the incident light would have registered on
the PM tube used to measure the reflected light. The cross-calibra
tion was determined by placing the PM tubes side by side facing the
UV beam. For some constant monochromatic incident flux the tubes were
moved in and out of the beam and the resulting outputs were divided.
The cross-calibration curve is shown in Figure 3 with the appropriate
error bars and the values are listed in Table 1. The cross-calibra
tion errors are essentially the total errors of the experiment and
are a result of nonuniformities in the UV beam and the PM photocathodes.
These nonuniformity problems were of great concern and are discussed
in Appendix 5.
Calculation of the expected signal levels showed that the PM
tube used to measure the incident radiation could be saturated while
the PM measuring the reflected radiation would have a low signal-to-
noise ratio. This problem is a direct result of the reflectance
characteristics of the frost. It is assumed that the frost is a
Lambert reflector and thus distributes the incident flux according
to the cosine law. The angular distribution of reflected radiation
from CO2 cryodeposits has been measured by Smith et al. (1969) and was
found to be essentially Lambertian. As positioned in the chamber,
the reflected light PM tube had a collecting solid angle approximately
-3
10 times that of a hemisphere and thus the intensity of radiation on
_3
the reflected light PM tube was % 10 that of the incident light PM
tube. This difficulty was overcome by operating the incident light

-17-
RM tube at a reduced gain compared to the reflected light PH tube.
The PM tube gain was controlled by varying the applied high voltage.
The reflected light PM tube was always kept at 2950 volts while the
incident light PM tube high voltage ranged from 1600 to 2800 volts.
Following this procedure of reducing the gain of one PM tube
meant that a gain-volt calibration for this particular tube had to be
established so that measurements taken at a reduced gain could be
accurately extrapolated to the gain at 2950 volts. This was easily
accomplished by parking the PM tube in the light beam and changing
the applied high voltage over a broad range (1600-2950 volts). Next,
all PM tube outputs at reduced voltages were divided into the output
at 2950 volts. This function is plotted in Figure 4 and tabulated in
Table 2. It was found that signals at 2950 volts could be readily
predicted within an accuracy of 1 percent by recording the signals
at reduced voltages. In most instances the voltage of the incident
light PM tube ranged between 2400 and 2800 volts, and only in cases
where large incident light levels were required to obtain a respectable
reflected light signal (i.e., in wavelength regions of strong frost
absorption) did the high voltage have to be reduced to 1600 to 1800
volts. For this case the error is somewhat worse ( +3 percent). The
- 1 percent accuracy in the 2400 to 2800 volt range is attributed to
the highly commendable performance of the EMR solar-blind PM tube.
This gain-volt calibration was checked from time to time during the
experimental period and was always found to display this remarkable
accuracy.

GAIN AT 2950 VOLTS -r GAIN AT X-VOLTS
-18-
HIGH VOLTAGE
Figure 4
The Gain-Volt Calibration of PM 9553

-19-
Table 2
Gain-
-Volts Calibration Values
X-High Voltage
Patio Gain @ 2950 volts
Gain @ x-volts
1700
144.44
1800
92.86
1900
57.52
2000
35.14
2100
22.81
2200
14.77
2300
10.00
2400
6.67
2500
4.64
2600
3.17
2700
2.28
2800
1.60
2900
1.16
2950
1.00

-20-
Since the reflected light PM tube was stationary at 10 from
normal incidence, the total hemispheric'] reflectance could only be
measured by replacing the unknown reflector (NH^ and H2O frosts) by a
o
diffusely reflecting standard. This was only done for 3000A radiation
to establish a hemispherical reflectance and then all the remaining
wavelengths were adjusted from a relative to an absolute reflectivity.
Initially magnesium oxide was chosen as a standard. Magnesium
ribbon was burned and the oxide smoke collected on aluminum or stain
less steel plates. However, after some laboratory use and a review of
the literature, it was clear that magnesium oxide has a number of un
desirable characteristics such as rapid aging, large thicknesses are
required for opacity, the powder is quite fragile, and an uncertainty
o
in the value of total reflectivity at 3000A (see Benford et al., 1948a
and 1948b). It was thus decided to replace the magnesium oxide with
a barium sulfate standard. In contrast to magnesium oxide, barium
sulfate has the desirable properties of small changes in reflectance
with age, it can be purchased commercially as a powder or as a paint
from Eastman Kodak, and the paint is fairly durable. The reflectiv
ities of aged BaSO^ paint, fresh BaSO^ paint, and BaSO^ powder that
was measured in this study are shown in Figure 25, The photometry of
o o
BaSO^ from 2000A to 8000A is discussed in detail by Billmeyer (1969),
Grum and Luckey (1968), and in Kodak publications No. JJ-31 and No.
JJ-32.
A summary of the photometric procedure is as follows (the inci
dent and reflected light PM tubes are denoted by their respective serial
numbers, 9553 and 6157):

-21-
1) Measure IQ (X) with PM 9553 at reduced
gain (x-volts).
2) Calculate the IG (X) which PM 9553 would
register at 2950 highvolts using the
volts-gain calibration curve.
3) Employ the cross calibration curve to
determine IQ (X) which PM 6157 would
have registered.
4) Measure Ir (X), the reflected light
signal, with PM 6157, and divided it
by Io (x) steP 3-
5) Adjust the relative reflectivity
(Ir t X]/Io [X]), to a hemispherical
reflectivity by comparison to BaS04
at 3000A.
E. Source Gases
Ultrahigh pure ammonia (99.999 percent pure) was purchased
commercially from Air Products and Chemicals Inc. and proved to be
sufficiently pure. The primary foreign gas in the UHP ammonia is
nitrogen which is noncondensible at 77 K and was thus pumped out by
the diffusion pump during the frost growth. The NH3 source bottle
was connected to a ballast chamber and subsequently bled into the
frost chamber through a needle valve.
Obtaining pure water vapor was somewhat more difficult than
NH3. A commercial still was used to produce "conductivity water"
which was collected in a glass vacuum trap. This trap was cleaned
with chromic acid and leached with water from the still (approximately
200 F). After sufficient leaching, the water was collected in the
trap and tested with a conductivity meter. A conductivity of 0.5 x
-6 -1 -1
10 ohm cm was set as an acceptable purity.

-22-
The most difficult impurity to remove from water is ammonia
absorbed in the form of NH^OH. If the concentration of NH^OH is very
small it is totally dissociated into NH^+ and OH ions which are the
ions measured with the conductivity meter. From order-of-magnitude
calculations it can be shown (see Appendix 4) that for a conductivity
of 0.5 x 10 6 ohm ^ cm ^ the NH^+ ion concentration is approximately
0.2 ppm. This level of ammonia contamination in the water is much too
small to be seen in the ^0 frost spectrum. It is possible that a
greater amount of NH3 was introduced into the H2O frost by out-
gassing of previously adsorbed NH^ (from the chamber walls). This
NH3 out-gassing could only have been the order of 10 ppm which would
still be undetectable in H£0 frost reflectivities.
Once the conductivity water was collected, the trap was immed
iately connected to the frost chamber inlet line and also to a mechan
ical vacuum pump line. This vacuum pump was activated so that the
hot water boiled under vacuum. This process removed most of the
nitrogen from the water and once the trap was valved off it contained
pure water and a small amount of N2. During the H2O frost growth the
trap became cold due to the latent heat of evaporation and had to be
warmed slightly in order to maintain the room temperature ^0 vapor
pressure of 20 mmHg, thus insuring a uniform flow rate of H20 vapor
into the frost chamber.

III. H2O and NH3 Frost UV Reflectivities
A. Frost Growth Procedures
The techniques of controlling the growth environment for both
NH3 and H2O were strictly a result of trial and error. The understand
ing of how to control effectively the closely coupled parameters of
latent heat, heat transfer characteristics of the frosts, vapor pres
sure-temperature relationships, and phases of the solids was soon found
to be more difficult than the recording of photometric data. It be
came increasingly obvious that the initial growing conditions (flow
rate, chamber pressure, substrate roughness) dictated to a large ex
tent the growth patterns for the remaining growth period. It was also
recognized that it would be difficult to define these initial condi
tions. Particularly for NH3, the establishment of whether the frost
was in an amorphous or a cubic phase, or a combination of these two,
was a major experimental problem.
From the works of Dressier and Schnepp (1960), Mauer et al.
(1972), and Black et al. (1958), the techniques for obtaining an
essentially complete amorphous phase are well established. The data
of Dressier and Schnepp show that the solid cubic ammonia begins to
o
absorb about 200A deeper into the UV than the amorphous solid (see
Figure 12). Early data taken herein always showed that both the
amorphous and cubic frosts absorb strongly between 220oX and 2000X.
-23-

-24-
The early procedure for obtaining the NH3 cubic structure was to keep
the substrate temperature between 150 K and 190 K and to have a
large growth rate. The reflectivities of these cubic NH3 frosts
always agreed with the amorphous frost reflectivity and thus some
question of the phase was evident. This problem is discussed later
after the NH3 data are presented.
Once the H2O frosts were grown, it was clear that the most con
vincing technique of assuring a particular phase was to grow the amor
phous form first at a very slow rate and, after taking reflectance
measurements, to anneal the amorphous form to a cubic form. The H2O
phase change was visually obvious but temperature and pressure data
were recorded to substantiate the change (see Figure 29).
Another frost growing technique established through experimenta
tion was whether the chamber should be closed off or whether it should
be pumped on with the six-inch diffusion pump during the growth period.
For the NH3 and H2O sources, foreign gases were of sufficient quantity
that after a three-hour growth period a sealed frost chamber would have
a buffer gas present which would considerably alter the growth condi
tions. The basic effect of a buffer (or noncondensible) gas during
frost growth is to favor the growth of any frost particles protruding
from the surface since these particles see a larger concentration to
the condensible gas. The obvious decision was to leave the vacuum
pump open to the chamber if the effects of a buffer gas were not de
sired.
For most of the latest NH^ and 1^0 frosts, the following growth
procedure was followed:

-25-
1) Frost chamber pumped to 10 torr (on
occasion the flanges were baked out at
T ~ 100 F).
2) Frost dewar cooled down to temperature.
3) Flow started with vacuum pump on chamber.
The chamber pressure was never above
10~3 torr during the growth period.
4) Stop flow and measure reflectance at
3000X. If it was comparable to the
BaSO^ reference, the frost was assumed
to be optically thick; if not, the flow
was turned back on.
The mass flow rate through the needle valve was never measured
since it was impossible to determine what percentage of NH3 or H2O
was being frozen on the cryosurface or pumped out by the diffusion
pump. Also, the area of the cryosurface was poorly defined so that
even if the mass flow rate was known a thickness or density measure
ment would have large errors. The chamber pressure was used as an
indicator of the flow rate.
B. NH^ Frost Results
The UV reflectivity measurements of NH^ cubic and amorphous
frosts are presented in Figures 5 through 10. Each figure caption gives
pertinent information about the growth conditions. In addition, re
flectivity data for each frost are shown in tabular form in Tables
3-5. Smooth curves were drawn through the data points listed in each
table. Where zeros appear in the tables no data were taken. For
comparison the gaseous NH3 data of Watanabe et al. (1953) and the
solid NH3 data of Dressier and Schnepp have been reproduced in Figures
11 and 12.

-26-
The most striking result was the reflectivity exceeding 100 per
cent for frosts optically thick at 3000&. There are two possible causes
for this result; one is a consistent error in photometry due to PM
calibration errors and the second is that the frosts were not Lambert
reflectors for radiation in the middle ultraviolet. It must also be
mentioned that the different frost thicknesses for each experiment
can introduce a maximum uncertainty of 5 percent for all wavelengths.
For the thicker frosts the PM tube monitoring the reflected light was
closer to the frosts and thus had a larger collecting solid angle
while the PM tube measuring the incident flux always collected the
total light on the frost. This effect can be seen at 3000^ where the
reflectivity differs from frost to frost.
As for the spectral variations in reflectivity the uncertainty
lies in the Lambert assumption or calibration errors. On review of
the cross-calibration curve it is clear that at best the spectral
variance of reflectivity could be flat or increasing toward shorter
wavelengths, corresponding to negative or positive calibration errors,
respectively. As discussed in Appendix 5, the errors in cross
calibration were found to stem from nonuniformities of the PM tube
photocathodes and little could be done to correct this problem; how
ever, it seems safe to conclude that the reflectivities of both NH-j
and 1^0 cubic frosts increased toward shorter wavelengths. Supporting
evidence for this conclusion is the fact that some of the amorphous
frosts were indeed found to be constant in spectral reflectivity in the
wavelength region of no absorption.

-27-
It is just as possible that the frosts do not diffusely reflect
as a Lambert surface for UV wavelengths and have a scattering phase
function somewhat characteristic of a Rayleigh scattering media. This
would result in a higher reflectivity near normal incidence than that
of a Lambert reflector (e.g., the BaSO^ reference). Unfortunately this
possibility of an unknown scattering phase function could not be ex
amined in this experiment since it would require recording the reflectiv
ity at all reflecting angles, a task beyond the capabilities of the
system.
For the wavelength region below 2400X the ammonia frosts were
easily divisible into three groups: Group 1: NH3 #9, 10, and 19 are
cubic frosts that were grown directly into the cubic phase, Group 2:
NH3 #11 and 17b are cubic frosts formed by warming the amorphous
deposits until they were within the phase change temperature range
(when data were taken the frosts were recooled to 77 K), and
Group 3: NH^ #12 and 17a are amorphous frosts.
Group 3 shows no structure other than a continuum type cutoff
from 2400& to 1950$ and then was black out to 1400& (at I6O0X the
light source had sufficient output so that some return light could be
measured and the result was a reflectivity less than 1 percent which
was termed "black").
The Groups 1 and 2 were both cubic NH3 frosts but Group 2
showed a reduction of 20 percent in reflectivity between 2400& and
o
2200A and then a decrease of only 10 percent in reflectivity for the
next lOoX prior to the sharp cutoff between 2100& and 2000X. Why this

-28-
sharp drop in reflectivity between 2300& and 2200A and a lesser decrease
between 2200& and 2100A occurred in NH3 cubic frosts formed by annealing
from the amorphous phase could not be explained but is thought to be
related to the percentage of cubic structure obtained by annealing as
opposed to growing the cubic structure directly. The physics of these
phase changes clearly warrants further attention and could be best
studied by x-ray diffraction techniques.
For NH^ #19, a cubic frost formed directly as a result of a
large latent heat of formation load, the reflectivity has been plotted
to exemplify the repeatability of the photometric data (see Figure 10).
Three complete wavelength scans of the frost were made at one-hour
intervals. The vacuum pump was open to the frost chamber for the
entire three-hour period and the frost was kept at 77 K. The uncer
tainty in reflectivity is approximately 1.5 percent for \ > 2400&
and approximately 3 percent for \ < 240oX.
If the solid NH3 absorption data of Dressier and Schnepp
(Figure 12) are examined it is immediately obvious that one would
anticipate a cubic frost to start abosrption about 200? deeper in the
UV than an amorphous frost. On the contrary the cubic NH3 is observed
to reflect only 80S farther into UV than the amorphous NH3 (see Figure
9) and in some cases the cutoffs are essentially identical (see Figures
6, 7, and 8) .
It was because of the cubic NH3 absorption cutoff that in the
early NH3 experiments the question of what phase was being examined
arose. Since the results of Dressier and Schnepp give cubic NH3
absorption data only at 1875&, 1775&, and 1500&, the absorption

-29-
coefficients longward of 1875& obtained by an extrapolation are un
certain. The cubic NH^ frosts grown by annealing the amorphous frosts
show higher reflectivities at shorter wavelengths than those formed
directly into the cubic phase. However, the cubic NH3 of Dressier and
Schnepp was formed at a high temperature, not by annealing of the
amorphous NH3 and once again there appears to be conflict.
This discrepancy is difficult to analyze since amorphous NH^
obtained by deposition at 77 K could quite possibly have contained
an unknown amount of cubic NH^ and vice versa. This reasoning follows
from the x-ray work of Mauer et al. (1972). At best it can only be
concluded that a large percentage of what was assumed to be amorphous
phase was indeed amorphous. The technique followed in this study for
preparing the separate phases is given by Mauer. The only absolute
assurance of phase is to examine the solid NH3 with x-ray patterns prior
to measuring the UV reflectivities. Unfortunately this is beyond the
scope of this research. All of these arguments are likewise applicable
to the H2O frosts.
Photographs taken during the growth of the NH3 frosts and after
annealing are shown in Figures 13 and 14. In all photographs the en
largement is a factor of two. Explanations of each are given in the
figure captions.
C. H2O Frost Results
The UV reflectivity measurements of H2O cubic and amorphous frosts
are presented in Figures 15 through 22. The H20 amorphous and cubic
results are presented separately in cases where the cubic frosts were

Figure 5. NH^ #9: Cubic Phase
NOTES
a) This frost was grown directly as solid cubic by deposition at 180 K.
b) Buffer gas was He at 1000 y Hg.
c) Chamber was closed during recooling to 77 K.
d) Photometric data were taken at 77 K frost temperature.
e) Photograph of this frost shown in Figures 13c and d.
f) Resolution is given at top of the figure.
g) Growth period 67 minutes.

REFLECTIVITY (%)
1400 1600 1800 2000 2200 2400 2600
WAVELENGTH (A)
2600
3000
t

Figure 6. NH^ #10: Cubic Phase
NOTES
a) Frost was formed by very radid deposition at 77 K. Result was cubic NH^.
b) No buffer gas present.
c) Chamber closed.
d) Growth period 43 minutes.
e) Visual inspection showed NH^ crystals and extreme uniformity across frost dewar.

130
120
110
100
90
80
70
60
50
40
30
20
10
0
14 00
1600 1800
2 000
2200
2400 2600
WAVELENGTH (A)
2800
3 000
-33

Figure 7. NH^ #11: Cubic Phase
NOTES
a) The phase was established by annealing the amorphous phase at ^ 190^K.
b) Frost recooled with chamber closed to 77 K.
c) No buffer gas.
d) Amorphous was formed by very slow deposition at 77 K.
e) Growth period 3 hours 26 minutes.
-6
f) Chamber pressure % 10 torr when photometric data were recorded.
g) Note shoulder in reflectivity at 2200A.

130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600
1800 2000 2200 2400
WAVELENGTH (A)
2600 2800
3000
I
LO
Ln
I

Figure 8. NH^ #12: Amorphous Phase
NOTES
a)
Amorphous phase formed by very
slow deposition at 77
b)
Growth period 4 hours
c)
No buffer gas.
| |
d)
Chamber open to vacuum pump.
e)
Visual inspection showed very
fine-grained texture.
f)
Photograph of this frost shown
in Figures 13a and b.

REFLECTIVITY (%)
1
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
&
*

Table 3
-38-
Reflectivity vs. Wavelength
for NH3 #9, 10, 11, and 12
WAVELENGTH
NH3 9
NH3 10
NH3 11
NH3 12
3000.
98.59
98.53
100.14
98.26
2900.
101.14
97.15
96.28
100.57
2800.
104.27
98.99
98.37
107.33
2700.
110.94
100.83
97.89
107.33
2600.
109.17
104.35
98.21
117.30
2500.
112.50
105.26
95.31
75.29
2475.
0.0
0.0
0.0
0.0
2450.
0.0
0.0
0.0
0.0
2425.
0.0
0.0
0.0
0. 0
2400.
111.52
106.18
100.79
80.28
2375.
107.41
0.0
0.0
0.0
2350.
0.0
0.0
94.99
0.0
2325.
0.0
0.0
0.0
0.0
2300.
0.0
102.66
90.96
113.21
2275.
97.22
0.0
81.30
0.0
2250.
94.86
98.23
76.64
0.0
2225.
81.93
95.01
69.23
0.0
2200.
88.98
95.01
69.55
105.73
2175.
85.65
87.97
66.98
0.0
2150.
8 3.50
84.91
66.33
97. 54
2125.
79.58
73.90
60.21
94.52
2100.
75.66
65.94
57.48
85.44
2075.
64.68
53.09
47.17
72. 98
2050.
40.38
35.95
30.27
66. 75
2025.
15.29
17.75
12.07
47. 53
2000.
9.02
11.47
8.05
27.95
1975.
5.88
8. 72
6.76
16.20
1950.
0.0
8.11
0.0
10.68

Figure 9. NH3 #17a and b: Amorphous and Cubic Phases
NOTES
NH3 #17a is amorphous phase. #17b is cubic phase obtained by annealing amorphous
frost at 190 K.
No buffer gas. Chamber open to vacuum pump during amorphous growth.
-4 -6
Chamber pressure 2x10 torr during deposition and 10 torr when photometric
data were taken.
Note shoulder in cubic frost reflectivity at 22008 and difference in reflectivity
short of 22008 between amorphous and cubic.
The reflectivity of the copper substrate used in some of the NH3 frosts is also
shown.

1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH ()
-Otr-

Table 4
41-
Reflectivity vs. Wavelength for
NH^ #17a and b.
WAVELENGTH
NH3 17A
NH3 17B
3000.
98.60
92. 96
2900.
108.07
103.58
2800.
119.52
112.88
2 700.
122.51
113.21
2600.
122.18
116.03
2500.
125.66
116.37
2475.
127.82
117.36
2450.
128.32
117.69
2425.
129.31
118.86
2400.
126.33
120.68
2375.
126.49
119.35
2350.
125.33
118.86
2325.
125.83
114.87
2300.
122.67
112.38
2275.
120.85
107.57
2250.
115.70
102.42
2225.
112.55
97. 94
2200.
103.25
96.28
2175.
90.30
95.12
2150.
71.55
93.46
2125.
51.96
84. 83
2100.
33.53
78.02
2075.
21.75
66.90
2050.
16.77
57.93
2025.
13.61
36.52
2000.
11.12
11.95
1975.
7.47
6.47
1950.
4.48
6.31

Figure 10. NH3 #19a, b, and c: Cubic Phase
NOTES
Frost grown by deposition at high flow rate in order to form cubic phas
NH3 #19a, b, and c are all the same frost which was held at 77 K and
scans were taken at one-hour intervals.
For this frost the flow rate was initially low and when flow was in
creased the transition to cubic phase was very obvious by visual
inspection.

130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)

Table 5
Reflectivity vs. Wavelength for
NH3 #19a, b, and c
WAVELENGTH
NH3 19A
NH3 19B
NH3 19C
3000.
98.43
96.54
98.43
2900.
104.12
102.23
105.54
2600.
115.34
116.92
118.03
2700.
118.50
118.18
120.24
2600.
120.55
122.61
126.24
2500.
124.35
126.08
127.35
2475.
0.0
124.98
130.03
2450.
125.45
124.03
0.0
2425.
0.0
0.0
0.0
2400.
126.24
125.77
130.67
2375.
124.03
0.0
0.0
2350.
124.50
125.29
0.0
2325.
123.71
0.0
0.0
2300.
120.24
119.61
122.45
2275.
116.92
115.97
120.71
2250.
114.71
111.71
112.34
2225.
110.13
106.97
110.60
2200.
105.54
104.75
105.54
2175.
94.48
93.69
95.59
2150.
85.16
86.11
84.85
2125.
71.26
69.99
73.79
2100.
58.78
60.04
58.93
2075.
45.98
46.61
48.19
2050.
29.23
28.12
31.76
2025.
10.90
14.22
15.96
2000.
6.48
10.74
12.48
1975.
5.85
9.95
11.85
1950.
2.05
10.43
12.48

Absorption Coefficient (cm
CM
a
o
oo
Wavelength (X)
i
oi
i
Figure 11. NH3 Gas Absorption Coefficients (from Watanabe et al., 1953)
Absorption Cross Section (xlO

MOLAR ABSORPTION COEFFICIENT (LITER MOLE CM
-46-
2200
2000 1800 1600
WAVELENGTH (l )
1400
Figure 12. NH^ Solid Absorption Coefficients (from
Dressier and Schnepp, 1960)

Figure 13- Photographs of Cubic and Amorphous NH^
Amorphous Phase: Note the fine-grain size and matty
appearance.
Amorphous Phase: Note the two small pimples in lower
right of photo. These are believed to be cubic NH^
forming on the amorphous NH^.
Cubic NH3: This cubic phase obtained by rapid deposi
tion at ^ 180 K. The crystals can be seen as white
specks.
Same as Photo C. Presented to show large grain size
compared to the amorphous.

-48-

Figure 14. Photographs of NH3 Frosts Having Various
Textures
A) What was believed to be an amorphous frost with cubic growth
overlaying. Note how much brighter the pimples of cubic
appear.
B) An NH3 cubic frost after annealing at 180 K. Note icy
appearance.
C) NH^ cubic frost showing large grains.
D) NH3 cubic formed by annealing a cubic NH3 frost similar
to that shown in Photo C. Note how annealing greatly
reduces the voids between the grains.

-50-

-51-
repeatedly annealed and photometric data recorded after each annealing,
e.g., H20 #13 and 14. Each figure caption gives pertinent information
to that particular frost. Reflectivity values are also listed in
tabular form; where zeros appear in the tables no data were taken.
Photographs of selected H20 frosts are shown in Figures 26, 27,
and 28 with captions giving the important details. All photographs are
twice the actual size. Since the frosts grown with no buffer gas were
difficult to photograph most of the photographs are of buffer-gas-frosts.
The latter showed substantial structure best described as a ball or
"cauliflower" appearance while frosts grown without a buffer gas
photographed as uniform white reflectors and little or no structure
could be seen in the final prints.
For comparison, the absorption coefficients of H20 vapor
(Watanabe et al., 1953) and H20 hexagonal and amorphous solids
(Dressier and Schnepp,I960) are reproduced in Figures 23 and 24.
A number of unexpected results were obtained from the H20 frosts.
Most important of these are the absorption features centered at 2200$,
2075$, and 1925$. These features are more pronounced in the cubic
H20 frosts but also appear in some of the amorphous frost reflectivi
ties, e.g., H2O #14. Since it was anticipated that both NH^ and 1^0
would absorb in a continuum fashion there was immediate speculation
that the three absorption features were a result of an unknown con
taminant. As reviewed in Appendix 4, the technique for preparing
the water was examined in detail and several alternate methods of pre
paration were employed. Results were always the same, i.e., the
absorption features appeared in each 1^0 cubic frost.

-52-
Next, attention was directed to the photometry but the features
in question did not appear in the reflectance measurements of BaSO^
which was measured out to 1800X (see Figure 25). It was thus con
cluded that the photometry was correct.
o o
The absorption at 1925A shows the greatest half width, about 100A,
and is also characteristic of the manner in which the NH3 frost absorbed
in the cutoff region. Since the majority of the NH-j experiments pre
ceded the H2O experiments it was suggested that the NH3 had been
adsorbed by the aluminum flanges of the frost chamber and subsequently
outgassed during the H2O frost growth.
To eliminate this possibility the vacuum chamber was baked out
-6
and a 10 torr vacuum held in the chamber for several days. During
the next H2O experiment (H2O #13) an amorphous frost was grown, data
were taken, and then the frost was annealed above 160 K to obtain the
cubic H2O (this was standard procedure). The reflectance measurements
of the cubic H2O were recorded and the annealing procedure repeated
once again. Finally a known amount of NH3 was deposited on top of the
H2O frost. The result, shown in Figure 19, was an increase in re-
o
flectance in the 1925A absorption. The amount of NH3 admitted was
approximately fifty times that calculated to be present as a contam
inant.
Although this test of NH3 contamination seemed to indicate that
NH3 was not causing the absorption features in the H2O frosts, the
possibility still existed that the contaminant causing the features
could only be influential if embedded with the lattice structure of
the H2O crystals. This reasoning led directly to the theory of solid

-53-
state physics and the phenomena of exciton absorption.
Before examining the possibility of explaining the 1^0 frost
absorption features with exciton theory, several other interesting
results from the H2O frost experiments are discussed.
As was observed for the NH3 frosts the reflectivity of the H2O
o
cubic frosts increased from 3000A toward shorter wavelengths and in
most cases exceeded 100 percent. The reasoning behind this result is
identical to that given in the discussion of the NH3 frost data and
thus will not be discussed further.
H2O #5 and 7 were frosts grown within a closed chamber and thus
had a buffer gas present during formation. For #5 the buffer gas (N2)
background pressure was about 10 P Hg and for #7 about 100 u Hg. Both
of these frosts exhibited the growth of balls ranging in sizes from
< 1mm to about 8mm in diameter. The effect of a buffer gas is not
only to increase the growth rate of those frost grains protruding from
the surface but also to increase the conductive heat load from the
chamber walls. It is clear from Figure 26 that the growing balls have
small heat transfer paths to the substrate and this, coupled with the
increased conductive heat load, fixed the temperature at the growth
sites above the range for amorphous 1^0. Accordingly, both traces
shown in Figures 15 and 16 are the reflectivities of cubic 1^0 even
though the substrate temperature was 77 K. In Figure 15 the trace
labeled cubic (a) was the first scan of the "ball" frost and the
cubic (b) trace was taken after annealing the frost at temperatures up
to 225 K.

-54-
Thus, the result of a buffer gas is to mix the amorphous and cubic
phases and in most instances grow a cubic 1^0 frost overlying an amor
phous H2O. From Figures 15 and 16 it is also clear that since anneal
ing of the buffer-gas-frosts increases the grain size, the shape of
the absorption feature at 1925& is related to the grain size of the
material. No means were available to determine an average grain size
during the experiments so at best it can be concluded that an increased
grain dimension also increases the strength and width of the absorption
features. This grain size effect was also observed by Kieffer (1968)
in the infrared region for CO2 and H2O frosts.
For H2O, #10, 12, 13, and 14 the six-inch diffusion pump was open
to the frost chamber during the growth period so that the amorphous
H2O was easily formed since all noncondensible (buffer) gases are
pumped from the chamber. The amorphous H2O appeared grey and very
fine-grained. It was difficult to obtain an optically thick amorphous
frost since this phase of solid H2O usually grew as a translucent ice.
o
The reflectivity was checked intermittently at 3000A during the growth
period and when it converged toward the BaSO^ reflectivity the growth
was stopped. Typical growth periods ranged from three to five hours.
The annealing procedure outlined previously was employed to
change the amorphous phase to the cubic phase. The H2O phase change
o
occurred at 150 K and was accompanied by a rise in chamber pressure
(see Figure 29). A temperature rise, of the substrate, at the phase
change requires an exothermic reaction within the frost. It thus
follows that the energy level of the amorphous H2O must be greater,
for a given temperature, than the energy level of the cubic H2O for
a release of heat to occur during the phase change. This further

-55-
implies that the vapor pressure of amorphous H2O should be greater
than that of cubic H2O for a given temperature. When the pressure
jump was observed, at the 150 K phase change, it was felt that this
could have been a result of the difference in vapor pressures for
amorphous and cubic H20. Unfortunately, different vapor pressures
could not be positively established for the following reasons:
a) The vapor pressure for the H2O cubic has been assumed to be the
same as for H2O hexagonal ice. The vapor pressure of H2O hexagonal ice
has been well established (vapor pressure data taken from the Handbook
of Chemistry and Physics, 44th Ed.). This assumption can be in error.
8 o
b) The vapor pressure of the H2O cubic is % 6x10 torr at 150 K while
the outgassing of the frost chamber, when closed, increased the chamber
_3
pressure to % 10 torr in the same time required to warm the frost
from 77 K to 150 K. Consequently, the H2O pressure measured with
the Baratron during the frost warm-up is an unknown partial pressure
over the background pressure caused by outgassing and thus the H2O
vapor pressure is only approximately determinable, c) During the
growth period some noncondensible gas is always trapped within the
frost. When the amorphous H2O molecules reorient and migrate during
the change to an ordered structure (cubic phase) the trapped gas is
released. This is probably the best explanation of the rise in chamber
pressure during the phase change.
Once the cubic H2O was well established by annealing, the frost
o
was recooled to 77 K and during this cool-down the diffusion pump
was opened to the chamber after the pressure had been reduced to
'v 500 y Hg by refreezing of H20 vapor. The refreezing was somewhat
of a problem since it could cause a growth of very fine grains overlaying

-56-
the annealed-large-grain frost. The pumping of vapors from the chamber
during the recooling of the frost was done to minimize this problem.
For those H2O frosts grown following the above techniques the
amorphous H2O reflectivities are consistently lower than the cubic
H2O reflectivities (see Figures 17, 18, 19, and 20). The cubic H2O
reflectivities increase 15 to 20 percent from 3000X to 2400&. This
-0.6
increase in reflectivity nicely fits a \ law. As discussed by
Van de Hulst (1957) the scattering by large spherical particles is
-a
explained by a X law with a <1.0. For a = 0.6 the monodispersed
particle size is found to be 1, 1 p Since the frost particles are
not spherical and little is known of the change in complex index of
refraction with wavelength for cubic H2O, the X ^law is only
illustrative.
No consistent reason can be found to explain the reflectivity
of the amorphous H2O frosts. This is expected since the opacity,
grain configuration, and optical constants for the amorphous phase
are poorly understood. The physical characteristics of the amorphous
phase for both NH3 and H2O clearly warrants future attention.
The effects of annealing the H2O frosts were studied in H2O #14
(Figure 21). The results show a decrease in reflectivity shortward of
o o
2500A after each annealing. The cutoff at 1800A is also influenced,
o
i.e., the annealed frosts have lower reflectivities between 1800A
and 1700A.
o
The possibility of attributing the absorption features at 2200A,
o o
2075A, 1925A to a photon-exciton interaction can only be discussed in
general terms since (to the knowledge of the author) experimental
research, into the physics of solid water, is inconclusive at the

-57-
present time. The initial suggestion that exciton absorption is caus
ing the structure in the reflectivity of solid 1^0 observed in this
work was made by Prinz (1972). An excellent review of exciton theory
is given by Knox (1963).
The first explanation of absorption features in perfect insulat
ing materials was given by Frankel, Peierls, and Wannier (ref. Knox,
1963) in the early 1930's. Since that time, exciton-phonon-photon
interaction has been established and many elements and organic com
pounds studied. The most simplified exciton energy state model is the
hydrogen-like model and is similar to the Rydberg series.
The features observed in the 1^0 frost were reviewed in light of
the hydrogenic type series and no definite conclusion can be made at
this time.

Figure 15. 1^0 #5a and b: Cubic Phase
NOTES
This frost was grown with a 10 y Hg buffer gas (N2).
Both traces are cubic H2O. Trace "b" is after frost was annealed
at T ''/225o K for about one-half hour.

130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600
WAVELENGTH (A)
2 8 00
3 000
I
Ul
vO
!

Table 6
-60-
Reflectivity vs. Wavelength for
H2O #5a and b
WAVELENGTH
HOH 5A
HOH 5B
3000.
98.79
89. 89
2900.
106.62
103.77
2600.
120.86
113.03
2700.
117.48
119.97
2600.
121.93
121.40
2500.
127.98
128.69
2400.
126.74
129.05
2300.
124.07
129.41
2250.
0.0
0.0
2200.
122.11
111.61
2150.
119.08
0.0
2100.
117.48
104.31
2C75.
0.0
0.0
2050.
0.0
100.57
2025.
0.0
77.43
2000.
102.53
77.25
1975.
0.0
0.0
1950.
92.56
62.63
1925.
0.0
0. 0
1900.
99.50
64. 97
1875.
97.19
67. 64
185C.
102.71
73.34
1825.
98.26
75. 12
1800.
95.41
73.34
1775.
80. 10
59.63
1750.
64.26
37. 74
1725.
40.23
14.06
1700.
17.62
3.38

Figure 16. 1^0 #7a and b: Cubic Phase
NOTES
This frost was grown with a 10 ^ Hg buffer gas (N2).
Both traces are cubic 1^0. Trace "b" is after annealing for one and one-half
hours at T <\,225 K. Note the increased difference in reflectivity of trace
a and b with annealing time. In 1^0 #5 the annealing was done for about one-
half hour.

REFLECTIVITY (%)
i
130
120
HO
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
Os
N>

Table 7
Reflectivity vs. Wavelength for
1^0 #7a and b
WAVELENGTH
HOH 7A
HOH 7B
3000.
97.01
76. 18
2900.
98.43
87.22
2800.
103.58
87.04
2700.
106.27
95.76
2600.
113.39
96.30
2500.
115.34
101.46
2400.
119.97
100.21
2300.
114.45
100.93
2250.
0.0
94. 34
2200.
112.67
83.66
2150.
106.27
77.25
2100.
104.66
74.23
2075.
0.0
73.69
2050.
100.21
71.02
2025.
102.17
68.71
2000.
93.63
58.56
1975.
87.93
47. 53
1950.
74.76
39.87
1925.
75.83
38.09
1900.
68.71
37.38
1875.
68.17
42.01
1650.
67.64
46. 99
1825.
63.55
40. 76
1800.
54.29
45.57
1775.
36.67
36.49
1750.
19.22
24.92
1725.
10.15
13.35
1700.
4.63
5.52

Figure 17. HO #10: (a), Amorphous Phase; (b), Cubic Phase
NOTES
a) Trace (a) is very fine-grained amorphous ^0.
b) Trace (b) is cubic frost annealed from amorphous frost.
c) The amorphous frost in this experiment may not have been optically thick. The
frost was grown to examine the absorption features at 2200A, 2075A, and 1925&.
d) Chamber was closed.
e) No buffer gas.
f) Amorphous growth period was 1.5 hours.
, -. 6 o
g) The X law is shown by triangles for X > 2200A.

130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH ()
ft
<7\
*

Table 8
-66-
Reflectivity vs. Wavelength for
1^0 #10a and b
WAVELENGTH
HOH 10A
HOH 10B
3000.
72.20
98.17
2900.
69.91
101.23
2800.
72.01
103.52
2700.
70.86
108.49
2600.
72.01
109.82
2500.
73.53
109.63
2400.
72.58
114.98
2300.
72.01
115.36
2250.
73.34
106.39
2200.
66.09
99.32
2150.
68.57
97.60
2100.
65.51
91.49
2075.
70.10
77.55
2050.
63.98
78.31
2025.
62.27
82. 51
2000.
63.22
74. 30
1975.
56.73
64.37
1950.
50.42
62.65
1925.
48.70
61.31
1900.
49.47
59. 78
1875.
47.94
63.22
1850.
46.60
67.61
1825.
46.22
68. 57
1800.
41.06
70.48
1775.
32.28
65. 51
1750.
18.14
56.34
1725.
10.70
40. 87
1700.
4.77
17.57

Figure 18. 1^0 #12: (a), Amorphous Phase; (b), Cubic Phase
NOTES
Frost growth and annealing the same as 1^0 #10.
The same procedures were followed as in H2O #10 to establish the repeatability
of observing the absorption features.
The absorption features were seen in the H2O amorphous. This may indicate that
some cubic structure was present.

REFLECTIVITY {%)
J
130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
-89-

Table 9
Wavelength vs. Reflectivity for
H20 //12a and b
WAVELENGTH
HOH 12A
HOH 12B
3000.
84.45
97. 52
2900.
91.16
102.86
2800.
84.62
104.75
2700.
80.67
106.30
2600.
80.15
111.97
2500.
80.32
114.21
2400.
82.56
117.48
2300.
84.62
113.69
2250.
82.90
112.14
2200.
78.95
105.61
2150.
81.01
108.70
2100.
81.18
105.61
2075.
79.46
102.68
2050.
80.15
104.58
2025.
77.92
96. 84
2000.
75.51
96.84
1975.
66.05
89.78
1950.
60.37
84.28
1925.
54.70
78.09
1900.
52.63
85.31
1675.
50.22
85.83
1850.
48.16
88.41
1825.
44.03
86. 52
1800.
37.15
83.76
1775.
26.66
68. 97
1750.
16.86
48. 16
1725.
8.43
23. 56
1700.
4.13
7.57

Figure 19. 1^0 #13A: Amorphous Phase
NOTES
a) This frost grown to examine absorption features in the cubic phase when
NH^ is added on top of H^O cubic.
b) The amorphous frost was grown and reflectivity measured to see if cubic
absorption feature would be present. Only the 1925 dip was seen.
c) Frost grown at very slow flow rate.
d) Growth period 3.5 hours.

130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600
WAVELENGTH ()
28 00
3000

Figure 20. H2O #13: (i), (c), Cubic Phase; (d) NH3 Added Over H2O Cubic
NOTES
a) Traces (b) and (c) are cubic H^O. Trace (c) was annealed a second time. Trace
(d) is reflectivity of 1^0 cubic with a NH^ concentration of approximately .003
mole/liter. The NH^ was added on top of 1^0 cubic.
b) Each annealing of frost (b) and (c) was about 45 minutes.
c) Note the increase in reflectivity when NH^ was added. This could be explained
by an increase in scattering by the fine-grained NH^ particles formed over
coarse 1^0 cubic grains.

130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
I
U>
§

-74-
Table 10
Reflectivity vs. Wavelength for
H20 #13 a, b, c, and d
1 HOH FROST 13
WAVELENGTH
A
B
C
D
3000.
80.49
97.86
99. 87
100.87
2900.
83.00
99.87
94.02
101.03
2800.
79.83
100.87
102.54
104.21
2700.
76.49
99.53
0.0
102.04
2600.
76.32
105.21
101.37
106.88
2500.
79.66
100.87
0.0
109.55
2400.
85.67
109.89
108.88
111.22
2300.
89.01
107.88
0.0
110.22
2250.
91.35
107.21
o.c
107.55
2200.
89.34
100.87
99.20
102.04
2150.
88.84
100.87
98.20
101.87
2100.
86.00
100.20
92.18
98. 53
2075.
86.18
96. 36
91.85
96. 36
2050.
84.33
98.20
86.67
90. 85
2025.
87.01
90.85
87.51
86.51
2000.
79.49
85.34
74.31
79. 83
1975.
75.98
76. 32
66.80
70.47
1950.
68.64
73.65
60.45
69.47
1925.
65.46
71.31
60.79
66.47
1900.
64.63
74.31
61.12
68.47
1875.
63.96
74.65
64. 13
69.47
1850.
62.12
78.66
70.64
74.15
1825.
62.12
77.32
74.31
75.98
1800.
53.94
76.65
71.31
74. 65
1775.
38.91
59.79
59.79
59.45
1750.
19.71
36.07
34.07
36. 74
1725.
9.52
14.03
11.02
13.69
1700.
4.68
3.67
3.67
3.51

Figure 21. 1^0 #14: (a), Amorphous Phase
NOTES
a) The amorphous H2O showed the absorption features as did H2O #12. Again this
is believed to be a result of some cubic H2O present (possibly overlaying the
H2O amorphous).
b) Since the reflectivity increased going from 3000& to 240oX, this is also evi
dence of some 1^0 cubic present.

130
120
110
100
90
80
70
60
80
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
f
O'
I

Figure 22. 1^0 //14: (b), (c), (d), Cubic Phase
NOTES
a) This frost was annealed three separate times after being transformed to the
cubic phase.
b) Each annealing was about 45 minutes at T ^ 225 K.
c) Note the respectable decrease in reflectivity for wavelength between 2100X
and 2500X, after each annealing.
d) Note also the difference in reflectivity cutoff after each annealing.

130
120
110
100
90
80
70
60
50
40
30
20
10
0

79
Table 11
Reflectivity vs. Wavelength for
i^O #14 a, b, c, and d
1 HOH FROST 14
WAVELENGTH
A
B
C
D
3000.
105.72
97.77
102.51
90.42
2900.
109.55
109.85
0.0
94. 86
2800.
109.55
108.63
113.83
110.77
2700.
110.92
115.51
0.0
113.07
2600.
115.21
116.59
115.82
116.59
2500.
114.75
122.71
0.0
118.27
2400.
119.34
122.55
120.56
118.73
2300.
119.49
120.41
117.81
115.36
2250.
118.12
121.79
114.14
108.17
2200.
110.62
111.84
112.61
103.73
2150.
112.45
116.74
112.00
104.04
2100.
111.69
115.97
108.48
99. 14
2075.
106.03
115.36
104.65
99.91
2050.
104.35
109.39
105.42
96. 85
2025.
108.78
106.18
96.08
93.64
2000.
104.50
101.90
94.09
86. 14
1975.
92.72
95.47
85.37
82. 16
1950.
88.43
93.94
85.68
74.20
1925.
85.22
89.05
80.94
74.51
1900.
84.00
92.72
8 2.47
72.67
1875.
84.00
89. 35
84.15
76.35
1850.
82.31
85.99
84.46
79. 10
1825.
84.15
87. 36
83.69
79.25
1800.
77.42
83.08
76.81
70.84
1775.
65.18
70.07
58.45
49. 57
1750.
43.30
46.66
29.99
23. 87
1725.
23.26
20.96
8.11
7. 04
1700.
8.41
6.27
3.06
2.75

Absorption Coefficient (cm
30
25
20
15
10
5
0
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
Wavelength (ft)
Figure 23. H^O Vapor Absorption Coefficients (from Watanabe et al., 1953)
i
oo
o
Absorption Cross Section (xl0-1cm'

MOLAR ABSORPTION COEFFICIENT (LITER MOLE CM
WAVELENGTH (&)
Figure 24. H_0 Solid- Hexagonal and Amorphous
Absorption Coefficients (from
Dressier and Schnepp, 1960).

130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800
2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
Figure 25. BaSO^ and Stainless Steel Substrate Reflectivities
i
00
ro
I

Figure 26. Photographs of H^O Cubic Frosts
Grown with a Buffer Gas
NOTES
A) This 1^0 frost was grown at a slow rate with a 1000 y buffer
gas (N^). The vacancy in the center righthand of the photo
was once occupied by a "ball." Note the manner in which
the small balls were packed around the missing ball of frost.
B) Same photographs as "A" except for exposure time.

-84-

Figure 27. Photographs of 1^0 Buffer-Gas-Frosts
and 1^0 Amorphous Frost
NOTES
A) H0 frost grown at high deposition rate with buffer gas of
1000 y Hg (N). The "ball" type frost did not develop
because of tne nonuniform deposition obtained by a high
inlet flow of ^0 vapor. This frost is cubic even though
deposition took place at 77 K (see text for further
discussion).
B) Same as frost "A"; photo was taken at a different location
of dewar.
C) H_0 amorphous frost. Note the fine-grain size. Although
this frost cracked and no photometric data were taken, it
depicts the optical thickness of an amorphous frost. The
lighting is from the top of the picture and in the crack
which protrudes from the dewar it is clear the light
penetrates the frost for approximately 1/16 inch. The
small "ball" in this photo is the cubic 1^0 forming over
the amorphous.
D) Same frost as in photo "C" but taken earlier in the growth
period.

-86-
I

Figure 28. "Ball" Frost Growth Sequence
NOTES
These photographs were taken at 30-minute intervals during the growth
period. The buffer gas is ^ at 500 jj. Note in photograph "A" the
underlying amorphous 1^0 formed at 77 K during start of flow.

-88-

28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13
o oo
H2O Phase Change Data (Temperatures, K, shown are vp
from substrate thermocouple and pressures are from
the MKS Baratron transducer in p Hg).
Figure 29.

IV. Conclusions
The reflectivities of NH3 and H2O frosts were measured from
o o
3000A to 1400A and the effects on reflectivity of grain growth,
cubic or amorphous phase, and buffer gas were examined.
The reflectivites of the NH3 frosts are above 90 percent long-
o 00
ward of 2300A and below 1 percent from 1950A to 1400A. The absorp-
o o
tion cutoff from 2200A to 1950A occurs at longer wavelengths than
those expected from the absorption coefficients of solid NH3 measured
by Dressier and Schnepp (1960).
The H2O frost reflectivity measurements are quite different
from anything anticipated. Based on absorption data of Dressier and
Schnepp, a H2O frost should not become "black" until approximately
o
1500A. The present study shows that if the H2O frosts are optically
thick at 300oX the reflectivity will decrease to less than 10 percent
o o
between 1800A and 1700A. This aspect is more important in relation
to the study of the Jovian planets, in particular Saturn's rings.
o
Since the solar radiation at 1800A is approximately an order of magnitude
o
greater than at 1500A, detection of a solid H2O absorption will be
easier than previously thought.
It was also discovered that the H2O cubic frosts contained
absorption features which are not seen in 1^0 vapor. The possibility
-90-

-91-
of these absorptions being caused by excitons was suggested.
The reflectivity of NH3 #10, a cubic frost, is plotted against
the UV albedo of Jupiter (from Anderson et al., 1969) in Figure 30.
The Jupiter data were normalized to the frost data at 2400$. Wave
length ranges labeled A, B, and C correspond to the first drop in the
o o
Jovian albedo from 2200A to 2000A, to the somewhat level region from
00 o
2000A to 1800A and to the sharp cutoff at 1800A, respectively.
Comparison of the NH3 frost data with the Jupiter albedo in the
three wavelength regions shows good agreement only in the region A.
However, as stated by Anderson et al. (1969), region A can also be
explained with gaseous NH3 absorption. The absorption bands of NH3
gas would not appear in the Jupiter albedo since the resolution of the
rocket data was too low. In short, both gaseous and solid NH3 could
explain the Jupiter albedo in the wavelength range A.
It is unfortunate that no frost data could be recorded below
1900$. Nevertheless, the frost data between 1900$ and 2000$ do have a
trend similar to that of the Jupiter albedo.
No comment can be made concerning the wavelength range C. It
should be mentioned, however, that the Jupiter albedo cutoff at
o
1800A is very similar to the H2O frost cutoff.
Finally, no comparison of the ^0 frost data to the albedo of
the Saturnian ring can be made at present. The only available Saturnian
UV data are that from the 0A0-2 (see Wallace et al., 1972) and these
data are an integrated ring plus disk albedo.

Figure 30. Comparison of Jovian UV Albedo to NH^
Frost Reflectivity.

130
120
110
100
90
80
70
60
SO
40
30
20
10
0
1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
VO
LJ
I

APPENDICES

Appendices Introduction
The intent of the following appendices is to present a detailed
description of the instrumentation and procedures employed to measure
the UV reflectivities of the NH^ and ^0 frosts. The creditability
of the final results of any experimental research always depends upon
the completeness to which the techniques employed are understood.
Some of the procedures and apparatuses established in the design
phase of the project were found to contain undesirable aspects after
the equipment had been constructed and tested. Some of these diffi
culties were correctable but more important is that all of the tech
niques which contained experimental error were recognized.
It is hoped that the following appendices contain sufficient
detail to allow an accurate judgment of the reliability of the results.
-95-

Appendix 1
Light Source
The design of the light source was centered around the Evenson
type microwave cavity (see Fehsenfeld et al., 1964). The cavity was
powered by a Scintillonics 2450 MHz microwave generator. The cavity
takes a 13mm Quartz tube in which the discharge gas is located. For
this source the discharge gas was at pressures ranging from 1000 y
Hg to 500 y Hg. The continuum and line flux-output of the light source
o
is shown in Figure 30. For wavelengths short of 1800A it was initially
hoped that Kr could be used as the discharge gas since it exhibits a
continuum output at these wavelengths. Later it was found that the
contamination by caused the Kr continuum to be swamped by intense
lines. This presented no problem since the H2O and NH3 frosts were
o
strongly absorbing short of 2000A and the intense lines of the H£
discharge were exactly what was needed in order to record any reflected
radiation.
The discharge tube was continuously fed with laboratory grade
H2 through a Veeco variable leak valve and pumped by a mechanical
vacuum pump. Some contamination by ^ was always present and resulted
00
in a series of lines between 3400A and 2900A as shown in Figure 30.
When the frost data were taken some readjustment of the monochromator
-96-

-97-
o o o
slits was required for 2900A < X < 3400A but from 2900A to about
o
1680A the H£ continuum output proved to be very convenient.
The Evenson cavity could be "dead tuned" by adjustment of a coup
ling slider and a tuning stub. The reflected and forward power was
measured by a power meter in the Scintillonics power supply and when
the voltage standing wave ratio (VSWR) was approximately ten the
discharge was ignited by a Tesla coil. Some readjustment of the
cavity was required after the discharge was active. Later in the ex
periment period it was found that the cavity could be left tuned for
the active discharge and still be started with the Tesla. The dis
charge ran at a typical input power of 70 watts. The VSWR was reduced
to almost unity by careful tuning.
The light source was positioned as close to the entrance slit of
the monochromator as possible and was separated from the monochromator
by a 1mm thick, 1/2" diameter Mg?2 window. This MgF2 window had to be
cleaned from time to time, however, the light source was used for
approximately fifty hours and the MgF2 window seems to have lost little
transmissivity.
Overall the light source performed very satisfactorily through
out the experimental period.

Figure 31. Light Source Output as a Function
of Wavelength
NOTES
a) The insert shows the light source output from 3400$ and the
main figure shows the output from 2400$ to 1400&.
b) Monochromator slits were 200 p x 200 p.
c) PM 9553 recorded the output. High voltage was 2800 volts.
d) discharge pressure was 350 p.
e) Microwave generator was set at 70 watts forward power. The
VSWR was 1.2.

2400 2200 2000 1800 1600 1400
o
Light Source Relative Energy
-MCO.p'CnO'^jOOvO O i N5 UJ-P'LnCT'"J

Appendix 2
Monochromator
The monochromator was purchased commerically from the McPherson
Instrument Corporation, Model 218. Features of this model are the
following:
1) .3 meter focal length.
2) 1200 grooves per mm snap-in type grating.
3) high speed F/5.3 exit beam.
4) operational to 1000& when evacuated.
5) independently adjustable slits.
6) resolution capability 6&.
7) scans to "o"-order.
8) optical arrangement shown below.

0
rA/rAAA/c£m 8*am
£~Xj7~ &JT4M
-100-

-101-
The McPherson was evacuated to 10 torr with a 2" Chevron
cryo-baffle and a 2" oil diffusion pump backed by a 15 cfm Duo-Seal
forepump. The forepump and diffusion pump were separated by a Veeco
coaxial foreline trap to prevent forepump oil from backstreaming into
the monochromator. A 2" air operated gate valve separated the
McPherson from the vacuum pump, and was controlled by a safety elec
trical shutdown circuit. This combination was arranged such that in
the event of a power failure the gate valve would close and the dif
fusion pump would automatically shut off.
Scattered light from the McPherson was checked by monitoring the
output radiation with a solar-blind photomultiplier tube (PM tube)
while scanning the monochromator to wavelengths outside the range of
radiant sensitivity of the PM tube. Some signal could be seen below
1400$ if the PM tube gain was at a maximum setting and if the exit and
entrance slits of the monochromator were set to 2000 p x 2000 p (the max
imum opening of the slits). However, the slits were usually 800 p x
800 y during the frost experiments. At this setting the scattered
light for X < 1400$ was never above 5 percent of the light output for
o
1400 < X < 3400A. It is felt that if the PM tubes could have been more
solar-blina, for example the rubidium telluride photocathode long
o
wavelength cutoff is 3200A while the PM tubes used herein were cesium
o
telluride with a long wavelength cutoff at 3400A, the scattered light
component would have been essentially zero.

Appendix 3
Frost Chamber
The basic layout of the 6" Pyrex cross used for a frost chamber
has been given in the main text and only a few additional points need
to be made.
The chamber outgassing rate was measured by closing the 6" gate
valve and recording the rise in chamber pressure with the MKS Baratron.
The chamber would come to about 4 y Hg in one hour starting from 10 ^
torr. This high outgassing rate was due to the necessary instrumentation
located inside. Even though all PM tube wires were Teflon insulated
and no materials other than metals were used inside the chamber, the
outgassing could not be further reduced. The chamber was leak-tested
with a He leak detector and no leaks were found. It was concluded
that outgassing was at fault. It was hoped that the outgassed con
stituents were noncondensible at 77 K and were removed by keeping the
chamber open to the vacuum pumps during the frost growth period.
An important component of the frost chamber was the aluminum
cold shield labeled part "J" in Figure 1. The shield was an aluminum
tube 5 1/2" in diameter and was attached to the frost dewar. Thermo
couple measurements at the front of the tube indicated that the tube
was cooled to ^ 100 K while closer to the dewar the temperature was even
lower. This shield served to carry a large portion of the thermal
-102-

-103-
conduction heat load from the chamber walls and thus the frost was
exposed to "wall temperatures" of 100 K instead of 300 K. The shield
also served to keep any frost fragments from falling into the vacuum
pump and to attach several additional components, e.g., the BaSO^
reference and an observation light.
Components of the vacuum system employed to pump the frost
chamber were the following:
a) Six-inch oil diffusion pumps manufactured by
Norton Vacuum Equipment Division. The oil
used was DC-705.
b) Six-inch Chevron Cryo-Baffle was the cold trap
for the diffusion pump.
c) An air-operated 6" gate valve separated the
diffusion pump and cold trap from the frost
chamber. This gate valve was controlled by
the power failure safety-shutoff system which
also controlled the gate valve on the mono
chromator vacuum system.
d) The diffusion pump was backed by a 15 cfm
Duo-Seal mechanical vacuum pump. This same
pump was connected via vacuum valves to the
diffusion pump for the McPherson monochromator.
The vacuum grease used on all parts was Apiezon L.

Appendix 4
Source Gases
The NH^ gas was purchased commercially from Air Products and
Chemicals, Inc. The ultra high pure (UHP) grade NH3 was of accept
able purity (99.999 percent). It must be remembered, however, that if
the frost chamber were filled to a pressure of one atmosphere with
the UHP NH3 and then the frost dewar cooled to 77 K, the residual
pressure, after all the NH3 had frozen out, would be roughly 7.6 y Hg.
During a frost growth the quantity of NH3 consumed is estimated to
have been between twenty and fifty times the chamber volume (STP).
Clearly if the chamber is closed during the frost growth period the
background pressure would become substantial. For this reason the
chamber was left open to the vacuum pump throughout the growth period.
Perfection of the techniques needed to obtain pure ^0 vapor
required some months' trial and error.
The first H2O system attempted was to collect a sizable volume
of distilled water in a reservoir and if this supply were found to be
of an acceptable purity no further distillation would be required.
This approach was soon found to be completely inadequate. The water
was in constant contact with the aluminum flanges of the reservoir
and no means were available to check the water purity once the
reservoir had been filled.
-104-

-105-
After a number of minor alterations the most favorable pro
cedure was the following: 1) a small glass vacuum trap was employed
to collect the distilled water. This trap was cleaned with chromic
acid and continuously leached with conductivity water taken directly
from the still. 2) The conductivity of the water was checked with a
Leeds and Northrup conductivity meter until the acceptable limit of
ion concentration was reached. 3) This conductivity water was immed
iately connected to the frost chamber inlet line and to an evacuation
line. While the water was still hot a vacuum was slowly drawn on the
glass reservoir until the water just began to boil. This vacuum boil
ing removed almost all of the N2 absorbed in the water. 4) Each H2O
frost was grown from a new batch of conductivity water.
By specifying the acceptable water purity at a conductivity of
-6 -1 -1
.5 x 10 ohm cm the concentration of contaminants can readily be
calculated. It is commonly known that ammonia is the most difficult
contaminant to remove from water and since the most pronounced
absorption feature in the 1^0 frosts resembled the NH3 cutoff spectrum
it was assumed NH3 was the contaminant of greatest concern.
In the field of electrochemistry it was first established by
Arrhenius that the percentage of ionic dissociation rapidly increases
with decreasing concentration. The concentrations of NH0 dealt with
3 )
in this study were very small and it seemed safe to assume that the
NH3 was entirely dissociated into NH^+ and OH ions. A measurement
+
of the NH^ ion concentration is thus a direct measure of the NH3
concentration in the H2O vapor contained in the reservoir.

106-
To determine the concentration of NH
conductivity is first defined, i.e.,
the equivalent
A K/C
-1 2 -1
where A is the equivalent conductivity [ohm cm equivalents ], K
is the specific conductivity [ohm ^ cm and C is the concentration
_3
[equivalents cm ]. The concentration is measured in equivalents,
where one equivalent is the weight of substance necessary to give one
mole of H or OH in a neutralization reaction. Put another way, the
equivalence is the formula weight divided by the valence. For example,
if the ions that neutralize a one molar solution of H+ have a valence
of two then only .5 moles need be added. If thevvalence is unity the
concentration is in equivalents per liter or simply moles per liter.
For large concentrations of NH^OH the equivalent conductivity
is listed in the Handbook of Chemistry and Physics. The values are:
A r~c
238.00 0.0000
9.66 0.1000
5.66 0.1732
3.10 0.3160
Since NH^OH is a weak electrolyte the equivalent conductivity
increases very rapidly for dilute mixtures. To calculate C a conser
vative estimate of 100 for A is taken and the specific conductivity
was experimentally measured to be .5 x 10~6 ohm-1 cm-1. Thus C = K/A =
-9
5. x 10 equivalents per liter or moles per liter. For NH^OH there
are 35 grams per mole so C = .2 x 10~6 gr/cc.
For a concentration of .2 ppm of NH^ it is highly unlikely that
any feature in the H20 frosts could be caused by NH3>

-107-
It was thus concluded that the absorption features observed in
the H2O frosts were indeed due to absorption by water molecules or
as suggested by excitons.
The equilibrium vapor pressures for NH^ and 1^0 were calculated
for temperatures below those for which empirical data are available
by the following equation:
log10 P - T + b
where T is the absolute temperature, P is the pressure in mmHg and
a and b are cons tants.
This equation was fitted to the empirical data and then the
vapor pressures were calculated for temperature down to 77 K. The
final form of the vapor pressure-temperature relationships for NH^
and H^O were:
NH3: log1() P = -16.302xl02 + 9>9974
H20: log10 P = :.26-660x10 +10>5510<

Appendix 5
Calibration of PM Tubes
The total errors in the refelctivity measurements are due to the
uncertainties in the cross-calibration of the PM tubes used to measure
the incident and reflected radiation. No two photoelectric detectors
will generate the same electrical output for a given incident radiation.
Each detector will have a spectral sensitivity characteristic of its
photocathode and electron amplification mechanism.
As shown in Figure 1, the two PM tubes were mounted inside the
frost chamber: one to monitor the incident radiation (PM 9553) and
one to monitor the reflected radiation (PM 6157). There were no
optical components between PM 9553 and the frost and likewise for
PM 6157.
To measure the frost reflectivity it is only necessary to know
the relative sensitivity of the PM tubes since, once the incident
radiation (IQ) is measured by PM 9553, this I0 can be adjusted to what
PM 6157 would have measured. A secondary standard (BaSO^) was employed
to determine the hemispherical reflectivity; therefore, it is not
required that the PM tubes be calibrated on an absolute base. The
relative sensitivity of the PM tubes or cross-calibration must be ex
perimentally determined even though the manufacturer often supplies
the quantum efficiency of each PM tube.
-108-

-109-
The relative sensitivity of the PM tubes was determined by
placing the tubes in the frost chamber side by side facing the UV
beam. Each tube was moved in and out of the beam and their outputs
were divided. This procedure was followed until the wavelength range
o o
1400A to 3400A was covered.
The cross-calibration was measured before the frosts were grown
and then some six months later the cross-calibration was measured
again. The second cross-calibration was in poor agreement with the
pre-cross-calibration. This discrepancy was of great concern since
all of the frost reflectivities had been calculated using the pre
cross-calibration.
It was soon recognized that only one procedure in the calibra
tion scheme had changed from the pre- to post-cross-calibration.
The PM 9553 had never been moved from its holder, since it faced the
UV beam during the calibration and frost experiments, but the PM 6157
had to be moved to the front flange of the frost chamber in order to
record the reflected radiation. During the post-cross-calibration
the PM 6157 was repositioned in the holder exactly as it was for the
pre-cross-calibration with the one exception of roll orientation.
Upon examining the effect of changing the PM 6157 in roll it
was found that a difference in signal output of up to 30 percent
resulted. It was suggested that if the UV beam and the PM photocathode
was nonuniform the result would be a change in PM tube output with a
change in roll position. To establish this fact, the UV beam was
stopped down such that a spot approximately 1/16" diameter was formed
on the face of the PM tube. The PM tube was then moved laterally so

-110-
Position of light spot on photocathode
Figure 32. Photomultiplier Photocathode
Nonuniformities
NOTES
Dimensions "A" and "B" correspond to the dis
tance traversed by the 1/16 inch diameter spot
of light across the face of each PM tube.

-111-
that the 1/16" diameter spot traversed its face. This was done for
both PM tubes and the results are shown in Figure 31. The distances
labeled "A" and "B" correspond to the distance the spot moved across
the face of each PM tube. If the PM photocathodes were uniform this
signal would be relatively constant as the spot traversed the face
of the tubes. As is clear from the figure this is not the case. The
fact that nonuniformities in the photocathodes of the PM tubes were
causing the errors in the cross-calibration had thus been established.
To determine the scatter in the cross-calibration values the
PM 9553 was left fixed in its holder (the same position as during the
frost experiments) while the PM 6157 was changed in roll at 45 inter
vals. The cross-calibration was measured for all eight positions of
the PM 6157. For the various positions there were some overall dif
ferences in absolute sensitivity but the important quantity is the
difference in relative "spectral" sensitivity. The eight cross
calibration curves were thus normalized at 2000X and averaged to estab
lish the final cross-calibration curve. This is shown in Figure 3
with the appropriate error bars.
All of the frost data were reduced again using the post-cross-
calibration. Prior to discovering the calibration problem the experi
mental error was approximately 2 percent; however, since no improve
ment of the UV beam and PM photocathode uniformity could be made the
maximum spectral error in the frost photometry was 5 percent.

I
BIBLIOGRAPHY
Anderson, R. C., J. G. Pipes, A. L. Broadfoot, and L0 Wallace, 1969:
Spectra of Venus and Jupiter from 1800A to 3200A J. of Atm. Sci.,
26, 874-888.
Anderson, R. C. and J. G. Pipes, 1971: Jovian Ultraviolet Reflectivity
Compared to Absorption by Solid Ammonia, J. of Atm. Sci., 28,
1086-7.
Benford, F., S. Schwarz, and G. P. Lloyd, 1948a: Coefficients of Re
flection in the Ultraviolet of Magnesium Carbonate and Oxide,
J. of the Optical Society of America, 38, 964-5.
Benford, F., G. P. Lloyd, and S. Schwarz, 1948b: Coefficients of Re
flection of Magnesium Oxide and Magnesium Carbonate, J. of the
Optical Society of America, 35, 445-447.
Billmeyer, F. W., Jr., 1969: Part X: White Reflectance Standards,
Optical Spectra, Jan/Feb.
Black, I. A., L. H. Bolz, F. P. Brooks, F. A. Mauer, and H. S. Peiser,
1958: A Liquid-Helium Cold Cell for Use with an X-ray Diffracto
meter, J. of Research of the National Bureau of Standards, 61,
367-371.
Dressier, K. and 0. Schnepp, 1960: Absorption Spectra of Solid Methone,
Ammonia and Ice in the Vacuum Ultraviolet, J. of Chemical Physics,
33, 270-274.
Eastman Kodak, 1969: 'Eastman White Reflectance Paint,' Eastman White
Reflectance Standard, #JJ-32 (paint) and JJ-31 (standard).
Fehsenfeld, F. C., K. M. Evenson, and H. P. Broida, 1964: NBS Report
#8701, Microwave Discharge Cavities Operating at 2450 MHZ.
Grum, F. and G. W. Luckey, 1968: Optical Sphere Paint and a Working
Standard of Reflectance, Applied Optics, _7, 2289-94.
Harrison, H. and R. I. Schoen, 1967: Evaporation of Ice in Space:
Saturn's Rings, Science, 157, 1175-6.
.-112-

-113-
Kief fer H. H., 1968: Near Infrared Spectral Reflectances of Simu
lated Martian Frosts, Dissertation, California Institute of
Technology.
, 1969: Reflectance Spectrometer/Environmental Chamber
for Frosts, Applied Optics, 2 2497-2500.
, 1970: Spectral Reflectance of CO2-H2O Frosts, J. of
Geophysical Research, 75, 501-9.
Knox, R. S., 1963: Theory of Excitons, Academic Press (Supplement 5).
Kuiper, G. P., D. P. Cruikshank, and U. Fink, 1970a: The Composition
of Saturn's Rings, Sky and Telescope, 39^, 14.
, 1970b: (letter to editor), Sky and Telescope, 22. 80.
Lewis, J. S., 1969: The Clouds of Jupiter and the NH3 -H2O and NH3 -H2S
System, Icarus, J^ 365-378.
Mauer, F. A., L. H. Bolz, H. S. Peiser, and H. F. McMurdie, 1972:
(private communication, Notes on non-cubic NH^).
Owen, T., 1965: Saturn's Ring and the Satellites of Jupiter: Inter
pretations of Infrared Spectra, Science, 149, 974-5.
Pilcher, C. G., C. R. Chapman, L. A. Lebofsky, and H. H. Kieffer,
1970: Saturn's Rings: Identification of Water Frost, Science,
167, 1372-3.
Prinz, G. A., 1972: (private communications), U. S. Naval Research
Laboratory.
Schnepp, 0. and K. Dressier, 1960: Absorption Spectra of Solid Xe, Hr,
and Ar in the Vacuum Ultraviolet, J. of Chemical Physics, 33,
49-55.
Seiber, B. A., B. E. Wood, A. M. Smith, P. R. Muller, 1970: Density of
Low Temperature Ice, Science, 170, 652-4.
Smith, A. M., K. E. Tempelmeyer, P. R. Muller, and B. E. Woods, 1969:
Angular Distribution of Visible and Near I. R. Radiation Re
flected from CO2 Cryodeposits, AIAA Journal, 7_, 2274-80.
Van De Hulst, H. C., 1957: Light Scattering by Small Particles, John
Wiley & Sons, Inc.

-114-
Wallace, L., J. J. Coldwell, and B. D. Savage, 1972: Ultraviolet
Photometry from the 0A0-1II Observations of Venus, Mars,
Jupiter, and Saturn Longward of 2000X, The Astrophysical J.,
172.
Watanabe, K., M. Zelikoff, and E. C. Y. Inn, 1953: AFCRC Tech. Rpt.
No. 53-23, Geophys. Res. Paper No. 21.
Wood, B. E. and A. M. Smith, 1968: Spectral Reflectance of Water and
Carbon Dioxide Cryodeposits from .36 to 1.15p, AIAA J. T_, 1362-
1367.
Wood, B. E., A. M. Smith, J. A. Roux, and B. A. Seiber, 1971:
Spectral Infrared Reflectance of HoO Condensed on LN2~Cooled
Surfaces in Vacuum, AIAA J., 9^, 1836-1842.

BIOGRAPHICAL SKETCH
John G. Pipes was bom on September 6, 1945, in Gardner, Massa
chusetts. He graduated from Dan McCarty High School, Ft. Pierce,
Florida, in June, 1963. In September, 1963, he entered Indian River
Junior College in Ft. Pierce and received the Associate of Arts degree
two years later. In September, 1965, he transferred to the University
of Florida, where he received the degree of Bachelor of Science in
Aerospace Engineering in December, 1967. In January, 1968, he enrolled
in the Graduate School of the University of Florida. He spent April
and May of 1968 as a visiting graduate student at Kitt Peak National
Observatory in Tuscon, Arizona during the preparation of an Aerobee
rocket package for the observation of the ultraviolet spectrum of
Jupiter.
In June of 1969 he received the Master of Science in aerospace
engineering, having done graduate work in planetary atmospheres and
engineering optics. From that date until the present he has pursued
the degree of Doctor of Philosophy at the University of Florida.
John G. Pipes is married to the former Betty Anne Veber and
has two children, Michael, age four, and Pamela, age two.
-115-

I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Roland C. Anderson, Chairman
Associate Professor of Aerospace
Engineering
I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy. /¡A
Alex G^ sWt
Professor of Physics and Astronomy
I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
> .
Alex E. S. Green
Graduate Research Professor in
Physics and Astronomy
I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Bernard M. Leadon
Professor of Aerospace Engineering
I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Tiic-u/ /. yy.il *-c=L i. tt J7
David T. Williams
Professor of Aerospace Engineering

I certify that I have read this study and that, in my opinion,
it conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.
Mark H. Clarkson
Professor of Aerospace Engineering
This dissertation was submitted to the Dean of the College of Engi
neering and to the Graduate Council, and was accepted as partial fulfill
ment of the requirements for the degree of Doctor of Philosophy.
June, 1972
Dean, Graduate School




Figure 6. NH^ #10: Cubic Phase
NOTES
a) Frost was formed by very radid deposition at 77 K. Result was cubic NH^.
b) No buffer gas present.
c) Chamber closed.
d) Growth period 43 minutes.
e) Visual inspection showed NH^ crystals and extreme uniformity across frost dewar.


Figure 5. NH^ #9: Cubic Phase
NOTES
a) This frost was grown directly as solid cubic by deposition at 180 K.
b) Buffer gas was He at 1000 y Hg.
c) Chamber was closed during recooling to 77 K.
d) Photometric data were taken at 77 K frost temperature.
e) Photograph of this frost shown in Figures 13c and d.
f) Resolution is given at top of the figure.
g) Growth period 67 minutes.


-24-
The early procedure for obtaining the NH3 cubic structure was to keep
the substrate temperature between 150 K and 190 K and to have a
large growth rate. The reflectivities of these cubic NH3 frosts
always agreed with the amorphous frost reflectivity and thus some
question of the phase was evident. This problem is discussed later
after the NH3 data are presented.
Once the H2O frosts were grown, it was clear that the most con
vincing technique of assuring a particular phase was to grow the amor
phous form first at a very slow rate and, after taking reflectance
measurements, to anneal the amorphous form to a cubic form. The H2O
phase change was visually obvious but temperature and pressure data
were recorded to substantiate the change (see Figure 29).
Another frost growing technique established through experimenta
tion was whether the chamber should be closed off or whether it should
be pumped on with the six-inch diffusion pump during the growth period.
For the NH3 and H2O sources, foreign gases were of sufficient quantity
that after a three-hour growth period a sealed frost chamber would have
a buffer gas present which would considerably alter the growth condi
tions. The basic effect of a buffer (or noncondensible) gas during
frost growth is to favor the growth of any frost particles protruding
from the surface since these particles see a larger concentration to
the condensible gas. The obvious decision was to leave the vacuum
pump open to the chamber if the effects of a buffer gas were not de
sired.
For most of the latest NH^ and 1^0 frosts, the following growth
procedure was followed:


Appendix 1
Light Source
The design of the light source was centered around the Evenson
type microwave cavity (see Fehsenfeld et al., 1964). The cavity was
powered by a Scintillonics 2450 MHz microwave generator. The cavity
takes a 13mm Quartz tube in which the discharge gas is located. For
this source the discharge gas was at pressures ranging from 1000 y
Hg to 500 y Hg. The continuum and line flux-output of the light source
o
is shown in Figure 30. For wavelengths short of 1800A it was initially
hoped that Kr could be used as the discharge gas since it exhibits a
continuum output at these wavelengths. Later it was found that the
contamination by caused the Kr continuum to be swamped by intense
lines. This presented no problem since the H2O and NH3 frosts were
o
strongly absorbing short of 2000A and the intense lines of the H£
discharge were exactly what was needed in order to record any reflected
radiation.
The discharge tube was continuously fed with laboratory grade
H2 through a Veeco variable leak valve and pumped by a mechanical
vacuum pump. Some contamination by ^ was always present and resulted
00
in a series of lines between 3400A and 2900A as shown in Figure 30.
When the frost data were taken some readjustment of the monochromator
-96-


-29-
coefficients longward of 1875& obtained by an extrapolation are un
certain. The cubic NH^ frosts grown by annealing the amorphous frosts
show higher reflectivities at shorter wavelengths than those formed
directly into the cubic phase. However, the cubic NH3 of Dressier and
Schnepp was formed at a high temperature, not by annealing of the
amorphous NH3 and once again there appears to be conflict.
This discrepancy is difficult to analyze since amorphous NH^
obtained by deposition at 77 K could quite possibly have contained
an unknown amount of cubic NH^ and vice versa. This reasoning follows
from the x-ray work of Mauer et al. (1972). At best it can only be
concluded that a large percentage of what was assumed to be amorphous
phase was indeed amorphous. The technique followed in this study for
preparing the separate phases is given by Mauer. The only absolute
assurance of phase is to examine the solid NH3 with x-ray patterns prior
to measuring the UV reflectivities. Unfortunately this is beyond the
scope of this research. All of these arguments are likewise applicable
to the H2O frosts.
Photographs taken during the growth of the NH3 frosts and after
annealing are shown in Figures 13 and 14. In all photographs the en
largement is a factor of two. Explanations of each are given in the
figure captions.
C. H2O Frost Results
The UV reflectivity measurements of H2O cubic and amorphous frosts
are presented in Figures 15 through 22. The H20 amorphous and cubic
results are presented separately in cases where the cubic frosts were


Table 3
-38-
Reflectivity vs. Wavelength
for NH3 #9, 10, 11, and 12
WAVELENGTH
NH3 9
NH3 10
NH3 11
NH3 12
3000.
98.59
98.53
100.14
98.26
2900.
101.14
97.15
96.28
100.57
2800.
104.27
98.99
98.37
107.33
2700.
110.94
100.83
97.89
107.33
2600.
109.17
104.35
98.21
117.30
2500.
112.50
105.26
95.31
75.29
2475.
0.0
0.0
0.0
0.0
2450.
0.0
0.0
0.0
0.0
2425.
0.0
0.0
0.0
0. 0
2400.
111.52
106.18
100.79
80.28
2375.
107.41
0.0
0.0
0.0
2350.
0.0
0.0
94.99
0.0
2325.
0.0
0.0
0.0
0.0
2300.
0.0
102.66
90.96
113.21
2275.
97.22
0.0
81.30
0.0
2250.
94.86
98.23
76.64
0.0
2225.
81.93
95.01
69.23
0.0
2200.
88.98
95.01
69.55
105.73
2175.
85.65
87.97
66.98
0.0
2150.
8 3.50
84.91
66.33
97. 54
2125.
79.58
73.90
60.21
94.52
2100.
75.66
65.94
57.48
85.44
2075.
64.68
53.09
47.17
72. 98
2050.
40.38
35.95
30.27
66. 75
2025.
15.29
17.75
12.07
47. 53
2000.
9.02
11.47
8.05
27.95
1975.
5.88
8. 72
6.76
16.20
1950.
0.0
8.11
0.0
10.68


Figure 31. Light Source Output as a Function
of Wavelength
NOTES
a) The insert shows the light source output from 3400$ and the
main figure shows the output from 2400$ to 1400&.
b) Monochromator slits were 200 p x 200 p.
c) PM 9553 recorded the output. High voltage was 2800 volts.
d) discharge pressure was 350 p.
e) Microwave generator was set at 70 watts forward power. The
VSWR was 1.2.


Figure 2. The Photographs of the Experimental
Arrangement and Light Source
A. Experimental Arrangement
Components labeled are: 1) McPherson monochromator. 2) MKS
Baratron pressure transducer. 3) Frost chamber. 4) Preampli
fier.
B. Light Source
Components labeled are: 1) H£ discharge. 2) Microware cavity.
3) Leak valve for inlet gas. 4) Outlet pipe to mechanical
vacuum pump.


-50-


Table 9
Wavelength vs. Reflectivity for
H20 //12a and b
WAVELENGTH
HOH 12A
HOH 12B
3000.
84.45
97. 52
2900.
91.16
102.86
2800.
84.62
104.75
2700.
80.67
106.30
2600.
80.15
111.97
2500.
80.32
114.21
2400.
82.56
117.48
2300.
84.62
113.69
2250.
82.90
112.14
2200.
78.95
105.61
2150.
81.01
108.70
2100.
81.18
105.61
2075.
79.46
102.68
2050.
80.15
104.58
2025.
77.92
96. 84
2000.
75.51
96.84
1975.
66.05
89.78
1950.
60.37
84.28
1925.
54.70
78.09
1900.
52.63
85.31
1675.
50.22
85.83
1850.
48.16
88.41
1825.
44.03
86. 52
1800.
37.15
83.76
1775.
26.66
68. 97
1750.
16.86
48. 16
1725.
8.43
23. 56
1700.
4.13
7.57


2400 2200 2000 1800 1600 1400
o
Light Source Relative Energy
-MCO.p'CnO'^jOOvO O i N5 UJ-P'LnCT'"J


-5-
C. Basic Results
The reflectivities of fourteen separate cubic and amorphous water
frosts were recorded from 3000X to 1400?. All amorphous ^0 frosts
were grown at LN2 temperatures (77 K) while the growth rate, concen
tration of buffer gas, and substrate roughness were varied. The amor
phous frosts appeared milky and very fine-grained. In almost every
case the reflectivity was approximately 20 to 30 percent lower than the
cubic structure frosts.
A cubic water frost was obtained by allowing an amorphous frost
to warm gradually T % 4 deg/min) until the change of phase occurred
A t
at 150 K. The phase change was always accompanied by an exothermic
reaction, a release of adsorbed noncondensible gases (the frost chamber
pressure usually increased the order of 10 y Hg over a background
pressure of 5 y Hg), and an obvious increase in visible reflectivity.
The H2O amorphous frost's reflectivity is relatively constant
from 3000X to 2200X at which point the reflectivity decreases 20 per
cent in the region from 2200^ to 180C)£ prior to the absorption cutoff
o
at 1750A. Cubic water frosts exhibit an increasing reflectance from
00 00
3000A to 2300A and then three absorption features at 2200A, 2075A, and
o
1925A. The absorption cutoff is approximately the same for both water
o o
phases (1800A to 1700A) with the reflectivity dropping below 1 percent
from 170oS to 140oX. No data were taken for the hexagonal structured
solid water.
A total of nineteen ammonia frosts were grown; however, the re
sults of the first four were inconclusive and only helped to establish
experimental procedures. The same techniques for forming the H20


-105-
After a number of minor alterations the most favorable pro
cedure was the following: 1) a small glass vacuum trap was employed
to collect the distilled water. This trap was cleaned with chromic
acid and continuously leached with conductivity water taken directly
from the still. 2) The conductivity of the water was checked with a
Leeds and Northrup conductivity meter until the acceptable limit of
ion concentration was reached. 3) This conductivity water was immed
iately connected to the frost chamber inlet line and to an evacuation
line. While the water was still hot a vacuum was slowly drawn on the
glass reservoir until the water just began to boil. This vacuum boil
ing removed almost all of the N2 absorbed in the water. 4) Each H2O
frost was grown from a new batch of conductivity water.
By specifying the acceptable water purity at a conductivity of
-6 -1 -1
.5 x 10 ohm cm the concentration of contaminants can readily be
calculated. It is commonly known that ammonia is the most difficult
contaminant to remove from water and since the most pronounced
absorption feature in the 1^0 frosts resembled the NH3 cutoff spectrum
it was assumed NH3 was the contaminant of greatest concern.
In the field of electrochemistry it was first established by
Arrhenius that the percentage of ionic dissociation rapidly increases
with decreasing concentration. The concentrations of NH0 dealt with
3 )
in this study were very small and it seemed safe to assume that the
NH3 was entirely dissociated into NH^+ and OH ions. A measurement
+
of the NH^ ion concentration is thus a direct measure of the NH3
concentration in the H2O vapor contained in the reservoir.


I. Introduction
A. Impetus
Middle ultraviolet spectra for most atoms and molecules in the
gaseous phase, at least those relevant to the atmospheric sciences,
have been recorded and in some instances data exist for the liquid
and solid phase. The objective of the research report here was to
determine the ultraviolet reflectivities of frosts composed of solid
ammonia and water. The optical properties of atmospheric gases in the
solid phase have been becoming increasingly important because of the
renewed interest in the Jovian planets which has been prompted by the
space program. Prime examples are the discussions by Pilcher et al.
(1970), Kuiper et al. (1970a), and Kuiper et al. (1970b) of the Saturn
ian ring systems, believed to be covered by either an ammonia or water
frost. These experimenters examined the near infrared and visible
region while the 0A0-2 (Wallace et al., 1972) recorded the Saturnian
o
UV reflectivity down to 2250A. The temperature of the Saturnian rings
is believed to be approximately 90 K (see Owen [1965], Harrison and
Schoen [1967]).
In the case of Jupiter, Lewis (1969) has generated atmospheric
models and concluded that ammonia ice clouds are present in the upper
regions of the planet's atmosphere. In point of fact, it was this
-1-


-17-
RM tube at a reduced gain compared to the reflected light PH tube.
The PM tube gain was controlled by varying the applied high voltage.
The reflected light PM tube was always kept at 2950 volts while the
incident light PM tube high voltage ranged from 1600 to 2800 volts.
Following this procedure of reducing the gain of one PM tube
meant that a gain-volt calibration for this particular tube had to be
established so that measurements taken at a reduced gain could be
accurately extrapolated to the gain at 2950 volts. This was easily
accomplished by parking the PM tube in the light beam and changing
the applied high voltage over a broad range (1600-2950 volts). Next,
all PM tube outputs at reduced voltages were divided into the output
at 2950 volts. This function is plotted in Figure 4 and tabulated in
Table 2. It was found that signals at 2950 volts could be readily
predicted within an accuracy of 1 percent by recording the signals
at reduced voltages. In most instances the voltage of the incident
light PM tube ranged between 2400 and 2800 volts, and only in cases
where large incident light levels were required to obtain a respectable
reflected light signal (i.e., in wavelength regions of strong frost
absorption) did the high voltage have to be reduced to 1600 to 1800
volts. For this case the error is somewhat worse ( +3 percent). The
- 1 percent accuracy in the 2400 to 2800 volt range is attributed to
the highly commendable performance of the EMR solar-blind PM tube.
This gain-volt calibration was checked from time to time during the
experimental period and was always found to display this remarkable
accuracy.


Appendices Introduction
The intent of the following appendices is to present a detailed
description of the instrumentation and procedures employed to measure
the UV reflectivities of the NH^ and ^0 frosts. The creditability
of the final results of any experimental research always depends upon
the completeness to which the techniques employed are understood.
Some of the procedures and apparatuses established in the design
phase of the project were found to contain undesirable aspects after
the equipment had been constructed and tested. Some of these diffi
culties were correctable but more important is that all of the tech
niques which contained experimental error were recognized.
It is hoped that the following appendices contain sufficient
detail to allow an accurate judgment of the reliability of the results.
-95-


130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
I
U>
§


-2-
very prediction of solid NH^ clouds in conjunction with the UV rocket
spectrum of Jupiter obtained by Anderson et al. (1969) that provided
the impetus for this study.
Anderson et al. (1969) could not explain the sharp cutoff of the
o
Jupiter albedo at 1800A using NH3 gas and attributed it to an unknown
absorber. Later, employing the absorption coefficient data of
Dressier and Schnepp (1960) for solid cubic ammonia, Anderson and Pipes
(1971) suggested the unknown Jovian constituent to be solid cubic
ammonia. Since the data of Dressier and Schnepp (1960) for solid cubic
ammonia consisted of only two data points in the wavelength region of
interest, it was evident more experimental work on NH3 solid was
necessary.
It was thus proposed to grow NH^ and f^O frosts at LN^ tempera
tures until they become optically thick for wavelengths near 3000^
and to measure their reflectivities as far into the ultraviolet as
experimentally possible. The apparatus design employed many of the
experimental techniques used by the following investigators: Schnepp
and Dressier (1960), studies of solid Xe, Kr, Ar; Kieffer (1968, 1969,
1970), spectral reflectance of CO2-H2O frosts; and Wood et al. (1968,
1971), infrared reflectance of condensed on L^-cooled surfaces.
The work of Kieffer is by far most pertinent to the understanding of
frosts, since the others were examining optical properties of micron
thin clear ices or at best milky ices. Nevertheless, all these pub
lications were extremely helpful in defining the experimental
techniques employed in this study.


Absorption Coefficient (cm
30
25
20
15
10
5
0
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
Wavelength (ft)
Figure 23. H^O Vapor Absorption Coefficients (from Watanabe et al., 1953)
i
oo
o
Absorption Cross Section (xl0-1cm'


-110-
Position of light spot on photocathode
Figure 32. Photomultiplier Photocathode
Nonuniformities
NOTES
Dimensions "A" and "B" correspond to the dis
tance traversed by the 1/16 inch diameter spot
of light across the face of each PM tube.


REFLECTIVITY (%)
i
130
120
HO
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
Os
N>


1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH ()
-Otr-


Figure 22. 1^0 //14: (b), (c), (d), Cubic Phase
NOTES
a) This frost was annealed three separate times after being transformed to the
cubic phase.
b) Each annealing was about 45 minutes at T ^ 225 K.
c) Note the respectable decrease in reflectivity for wavelength between 2100X
and 2500X, after each annealing.
d) Note also the difference in reflectivity cutoff after each annealing.


Figure 1. The Schematic Diagram of the
Experimental Arrangement
A) Top Flange: Components: 1) high voltage and
PM tube electrical feedthrough. 2) PM tube
remote control. 3) thermocouple and ioniza
tion gauge.
B) Front Flange: Components: 1) all gas inlet
pipes. 2) holder for PM tube monitoring re
flected radiation. 3) observation port.
A) outlet to MRS Baratron pressure transducer.
C) Back Flange: Components: 1) LN2 inlet and
outlet feedthroughs. 2) temperature thermo
couple feedthrough. 3) roughing line vacuum
valve.
D) Six inch air operated gate valve.
E) Six inch chevron cryo-baffle.
F) Six inch oil diffusion pump.
G) Cryo-surface
H) PM tube monitoring incident radiation.
I) PM tube monitoring reflected radiation.
J) Aluminum cold shiled.
K) Monochromatic incident radiation.
L) Frost and reflected radiation
M) LN2 reservoir.
N) LN2 inlet.
O) LN2 feedthroughs.
P) Connection flange containing pin hole and
MgF£ window.
Q) McPherson monochromator.
R) Two inch chevron cryo-baffle.
S) Two inch oil diffusion pump.
T) Outlet to mechanical vacuum pump
U) One-half inch aluminum base plate.


-4-
cannot be obtained again by cooling the cubic to temperatures as low
as LHe.
An important aspect of frost spectroscopy is the characteristic
equilibrium vapor pressure. Since the frosts are grown and examined
-4 -6
in an evacuated chamber (typically 10 to 10 torr), it is essential
that their vapor pressures at temperatures is so low that absorp
tion by vapor is insignificant. No experimental vapor pressure data
exist for NH3 and H2O at 77 K; however, calculated vapor pressures
-25 -12
(see Appendix 4) are 10 torr for H2O and 10 torr for NH3. Thus,
the effect of gaseous absorption is unimportant.
Still another important consideration is the method of forming
the frost. Vapor deposition on a cryogenic surface is classified as
substrate cooling and is quite different from one of nature's prime
cooling mechanisms, i.e., radiative cooling. In the laboratory the
radiation is always a heat load on the frost instead of a heat loss;
however, it is essentially impossible to cool every black or grey
body surrounding the frost to temperatures lower than the frost. The
conductive heat load must also be considered and is no doubt much larger
than the radiative load even when the frost chamber is evacuated to
-6
10 torr. This of course is assuming the walls of the chamber to
be at room temperature. In short it would be difficult to simulate
even approximately the frosts that exist in nature (e.g., Saturn's rings
and the Martian polar cap); however, it is felt that valuable informa
tion can be extracted by growing frosts using substrate cooling.


Figure 14. Photographs of NH3 Frosts Having Various
Textures
A) What was believed to be an amorphous frost with cubic growth
overlaying. Note how much brighter the pimples of cubic
appear.
B) An NH3 cubic frost after annealing at 180 K. Note icy
appearance.
C) NH^ cubic frost showing large grains.
D) NH3 cubic formed by annealing a cubic NH3 frost similar
to that shown in Photo C. Note how annealing greatly
reduces the voids between the grains.


PM 6157 -7- PM 9553
m i i i i
6.6
6.4
6.2
6.0
5.9
5.6
5.4
5.2
5.0
I I I I I I I I I
J I I I I i I I I I I I I I I I
1600 1800 2000 2200 2400 2600
WAVELENGTH ()
2800
3000
Figure 3. The Cross-Calibration Curve. FM 6157 is the reflected light detector
and PM 9553 is the incident light detector.
i
-14


Figure 9. NH3 #17a and b: Amorphous and Cubic Phases
NOTES
NH3 #17a is amorphous phase. #17b is cubic phase obtained by annealing amorphous
frost at 190 K.
No buffer gas. Chamber open to vacuum pump during amorphous growth.
-4 -6
Chamber pressure 2x10 torr during deposition and 10 torr when photometric
data were taken.
Note shoulder in cubic frost reflectivity at 22008 and difference in reflectivity
short of 22008 between amorphous and cubic.
The reflectivity of the copper substrate used in some of the NH3 frosts is also
shown.


II. Experimentation
The experimental arrangement is shown schematically in Figure 1
and photographically in Figure 2. The experimentation is best de
scribed if subdivided into the following categories: A) the H£ dis
charge light source, B) McPherson monochromator, C) frost chamber and
cryosurface, D) Photometry, E) source gases.
A. Light Source
The light source was a flow-through electrodeless discharge type.
Hydrogen was used as a discharge gas at pressures between 500 and
1000 y Hg. Ifydrogen exhibits a uniform continuum from 3000^ to ap
proximately 1650& so that little or no readjustment of the monochrom
ator slits or photomultiplier tube gain was required. The source was
placed sufficiently close to the entrance slit so that the optics of
the monochromator were overfilled. The high temperature discharge
gas was separated from the monochromator vacuum by a MgF£ window. The
source proved to be quite flexible, contamination free, and extremely
stable for long periods of time. The spectral distribution of the
light source is given in Appendix 1.
B. Monochromator
A 0.3m scanning McPherson monochromator was employed. The
monochromatic energy requirements (never greater than 10~10watts/cm2-#)
-7-


Figure 27. Photographs of 1^0 Buffer-Gas-Frosts
and 1^0 Amorphous Frost
NOTES
A) H0 frost grown at high deposition rate with buffer gas of
1000 y Hg (N). The "ball" type frost did not develop
because of tne nonuniform deposition obtained by a high
inlet flow of ^0 vapor. This frost is cubic even though
deposition took place at 77 K (see text for further
discussion).
B) Same as frost "A"; photo was taken at a different location
of dewar.
C) H_0 amorphous frost. Note the fine-grain size. Although
this frost cracked and no photometric data were taken, it
depicts the optical thickness of an amorphous frost. The
lighting is from the top of the picture and in the crack
which protrudes from the dewar it is clear the light
penetrates the frost for approximately 1/16 inch. The
small "ball" in this photo is the cubic 1^0 forming over
the amorphous.
D) Same frost as in photo "C" but taken earlier in the growth
period.


130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH ()
ft
<7\
*


LIST OF FIGURES
Figure Page
1. The Schematic Diagram of the Experimental
Arrangement 9
2. The Photographs of the Experimental Arrangement
and Light Source 11
3. The Cross-Calibration Curve 14
4. The Gain-Volt Calibration of PM 9553 18
5. NH3 #9: Cubic Phase 31
6. NH3 #10: Cubic Phase 33
7. NH3 #11: Cubic Phase 35
8. NH3 #12: Amorphous Phase 37
9. NH3 #17a and b: Amorphous and Cubic Phases 40
10. NH3 #19a, b, and c: Cubic Phase 43
11. NH3 Gas Absorption Coefficients 45
12. NH3 Solid Absorption Coefficients 46
13. Photographs of Cubic and Amorphous NH3 48
14. Photographs of NH3 Frosts Having Various Textures 50
15. H2O #5a and b: Cubic Phase 59
16. H20 #7a and b: Cubic Phase 62
17. H20 #10: (a), Amorphous Phase; (b), Cubic Phase 65
18. H20 #12: (a), Amorphous Phase; (b), Cubic Phase 68
19. H20 #13A: (a), Amorphous Phase 71
vi


-55-
implies that the vapor pressure of amorphous H2O should be greater
than that of cubic H2O for a given temperature. When the pressure
jump was observed, at the 150 K phase change, it was felt that this
could have been a result of the difference in vapor pressures for
amorphous and cubic H20. Unfortunately, different vapor pressures
could not be positively established for the following reasons:
a) The vapor pressure for the H2O cubic has been assumed to be the
same as for H2O hexagonal ice. The vapor pressure of H2O hexagonal ice
has been well established (vapor pressure data taken from the Handbook
of Chemistry and Physics, 44th Ed.). This assumption can be in error.
8 o
b) The vapor pressure of the H2O cubic is % 6x10 torr at 150 K while
the outgassing of the frost chamber, when closed, increased the chamber
_3
pressure to % 10 torr in the same time required to warm the frost
from 77 K to 150 K. Consequently, the H2O pressure measured with
the Baratron during the frost warm-up is an unknown partial pressure
over the background pressure caused by outgassing and thus the H2O
vapor pressure is only approximately determinable, c) During the
growth period some noncondensible gas is always trapped within the
frost. When the amorphous H2O molecules reorient and migrate during
the change to an ordered structure (cubic phase) the trapped gas is
released. This is probably the best explanation of the rise in chamber
pressure during the phase change.
Once the cubic H2O was well established by annealing, the frost
o
was recooled to 77 K and during this cool-down the diffusion pump
was opened to the chamber after the pressure had been reduced to
'v 500 y Hg by refreezing of H20 vapor. The refreezing was somewhat
of a problem since it could cause a growth of very fine grains overlaying


Figure 26. Photographs of H^O Cubic Frosts
Grown with a Buffer Gas
NOTES
A) This 1^0 frost was grown at a slow rate with a 1000 y buffer
gas (N^). The vacancy in the center righthand of the photo
was once occupied by a "ball." Note the manner in which
the small balls were packed around the missing ball of frost.
B) Same photographs as "A" except for exposure time.


Table 5
Reflectivity vs. Wavelength for
NH3 #19a, b, and c
WAVELENGTH
NH3 19A
NH3 19B
NH3 19C
3000.
98.43
96.54
98.43
2900.
104.12
102.23
105.54
2600.
115.34
116.92
118.03
2700.
118.50
118.18
120.24
2600.
120.55
122.61
126.24
2500.
124.35
126.08
127.35
2475.
0.0
124.98
130.03
2450.
125.45
124.03
0.0
2425.
0.0
0.0
0.0
2400.
126.24
125.77
130.67
2375.
124.03
0.0
0.0
2350.
124.50
125.29
0.0
2325.
123.71
0.0
0.0
2300.
120.24
119.61
122.45
2275.
116.92
115.97
120.71
2250.
114.71
111.71
112.34
2225.
110.13
106.97
110.60
2200.
105.54
104.75
105.54
2175.
94.48
93.69
95.59
2150.
85.16
86.11
84.85
2125.
71.26
69.99
73.79
2100.
58.78
60.04
58.93
2075.
45.98
46.61
48.19
2050.
29.23
28.12
31.76
2025.
10.90
14.22
15.96
2000.
6.48
10.74
12.48
1975.
5.85
9.95
11.85
1950.
2.05
10.43
12.48


Figure 21. 1^0 #14: (a), Amorphous Phase
NOTES
a) The amorphous H2O showed the absorption features as did H2O #12. Again this
is believed to be a result of some cubic H2O present (possibly overlaying the
H2O amorphous).
b) Since the reflectivity increased going from 3000& to 240oX, this is also evi
dence of some 1^0 cubic present.


-57-
present time. The initial suggestion that exciton absorption is caus
ing the structure in the reflectivity of solid 1^0 observed in this
work was made by Prinz (1972). An excellent review of exciton theory
is given by Knox (1963).
The first explanation of absorption features in perfect insulat
ing materials was given by Frankel, Peierls, and Wannier (ref. Knox,
1963) in the early 1930's. Since that time, exciton-phonon-photon
interaction has been established and many elements and organic com
pounds studied. The most simplified exciton energy state model is the
hydrogen-like model and is similar to the Rydberg series.
The features observed in the 1^0 frost were reviewed in light of
the hydrogenic type series and no definite conclusion can be made at
this time.


Appendix 2
Monochromator
The monochromator was purchased commerically from the McPherson
Instrument Corporation, Model 218. Features of this model are the
following:
1) .3 meter focal length.
2) 1200 grooves per mm snap-in type grating.
3) high speed F/5.3 exit beam.
4) operational to 1000& when evacuated.
5) independently adjustable slits.
6) resolution capability 6&.
7) scans to "o"-order.
8) optical arrangement shown below.

0
rA/rAAA/c£m 8*am
£~Xj7~ &JT4M
-100-


130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600
1800 2000 2200 2400
WAVELENGTH (A)
2600 2800
3000
I
LO
Ln
I


-56-
the annealed-large-grain frost. The pumping of vapors from the chamber
during the recooling of the frost was done to minimize this problem.
For those H2O frosts grown following the above techniques the
amorphous H2O reflectivities are consistently lower than the cubic
H2O reflectivities (see Figures 17, 18, 19, and 20). The cubic H2O
reflectivities increase 15 to 20 percent from 3000X to 2400&. This
-0.6
increase in reflectivity nicely fits a \ law. As discussed by
Van de Hulst (1957) the scattering by large spherical particles is
-a
explained by a X law with a <1.0. For a = 0.6 the monodispersed
particle size is found to be 1, 1 p Since the frost particles are
not spherical and little is known of the change in complex index of
refraction with wavelength for cubic H2O, the X ^law is only
illustrative.
No consistent reason can be found to explain the reflectivity
of the amorphous H2O frosts. This is expected since the opacity,
grain configuration, and optical constants for the amorphous phase
are poorly understood. The physical characteristics of the amorphous
phase for both NH3 and H2O clearly warrants future attention.
The effects of annealing the H2O frosts were studied in H2O #14
(Figure 21). The results show a decrease in reflectivity shortward of
o o
2500A after each annealing. The cutoff at 1800A is also influenced,
o
i.e., the annealed frosts have lower reflectivities between 1800A
and 1700A.
o
The possibility of attributing the absorption features at 2200A,
o o
2075A, 1925A to a photon-exciton interaction can only be discussed in
general terms since (to the knowledge of the author) experimental
research, into the physics of solid water, is inconclusive at the


GAIN AT 2950 VOLTS -r GAIN AT X-VOLTS
-18-
HIGH VOLTAGE
Figure 4
The Gain-Volt Calibration of PM 9553


-53-
state physics and the phenomena of exciton absorption.
Before examining the possibility of explaining the 1^0 frost
absorption features with exciton theory, several other interesting
results from the H2O frost experiments are discussed.
As was observed for the NH3 frosts the reflectivity of the H2O
o
cubic frosts increased from 3000A toward shorter wavelengths and in
most cases exceeded 100 percent. The reasoning behind this result is
identical to that given in the discussion of the NH3 frost data and
thus will not be discussed further.
H2O #5 and 7 were frosts grown within a closed chamber and thus
had a buffer gas present during formation. For #5 the buffer gas (N2)
background pressure was about 10 P Hg and for #7 about 100 u Hg. Both
of these frosts exhibited the growth of balls ranging in sizes from
< 1mm to about 8mm in diameter. The effect of a buffer gas is not
only to increase the growth rate of those frost grains protruding from
the surface but also to increase the conductive heat load from the
chamber walls. It is clear from Figure 26 that the growing balls have
small heat transfer paths to the substrate and this, coupled with the
increased conductive heat load, fixed the temperature at the growth
sites above the range for amorphous 1^0. Accordingly, both traces
shown in Figures 15 and 16 are the reflectivities of cubic 1^0 even
though the substrate temperature was 77 K. In Figure 15 the trace
labeled cubic (a) was the first scan of the "ball" frost and the
cubic (b) trace was taken after annealing the frost at temperatures up
to 225 K.


-97-
o o o
slits was required for 2900A < X < 3400A but from 2900A to about
o
1680A the H£ continuum output proved to be very convenient.
The Evenson cavity could be "dead tuned" by adjustment of a coup
ling slider and a tuning stub. The reflected and forward power was
measured by a power meter in the Scintillonics power supply and when
the voltage standing wave ratio (VSWR) was approximately ten the
discharge was ignited by a Tesla coil. Some readjustment of the
cavity was required after the discharge was active. Later in the ex
periment period it was found that the cavity could be left tuned for
the active discharge and still be started with the Tesla. The dis
charge ran at a typical input power of 70 watts. The VSWR was reduced
to almost unity by careful tuning.
The light source was positioned as close to the entrance slit of
the monochromator as possible and was separated from the monochromator
by a 1mm thick, 1/2" diameter Mg?2 window. This MgF2 window had to be
cleaned from time to time, however, the light source was used for
approximately fifty hours and the MgF2 window seems to have lost little
transmissivity.
Overall the light source performed very satisfactorily through
out the experimental period.


APPENDICES




-88-


Figure 8. NH^ #12: Amorphous Phase
NOTES
a)
Amorphous phase formed by very
slow deposition at 77
b)
Growth period 4 hours
c)
No buffer gas.
| |
d)
Chamber open to vacuum pump.
e)
Visual inspection showed very
fine-grained texture.
f)
Photograph of this frost shown
in Figures 13a and b.


28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13
o oo
H2O Phase Change Data (Temperatures, K, shown are vp
from substrate thermocouple and pressures are from
the MKS Baratron transducer in p Hg).
Figure 29.


UV REFLECTANCE OF FROSTS
COMPOSED OF WATER AND AMMONIA
By
John Gilbert Pipes
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1972


Appendix 4
Source Gases
The NH^ gas was purchased commercially from Air Products and
Chemicals, Inc. The ultra high pure (UHP) grade NH3 was of accept
able purity (99.999 percent). It must be remembered, however, that if
the frost chamber were filled to a pressure of one atmosphere with
the UHP NH3 and then the frost dewar cooled to 77 K, the residual
pressure, after all the NH3 had frozen out, would be roughly 7.6 y Hg.
During a frost growth the quantity of NH3 consumed is estimated to
have been between twenty and fifty times the chamber volume (STP).
Clearly if the chamber is closed during the frost growth period the
background pressure would become substantial. For this reason the
chamber was left open to the vacuum pump throughout the growth period.
Perfection of the techniques needed to obtain pure ^0 vapor
required some months' trial and error.
The first H2O system attempted was to collect a sizable volume
of distilled water in a reservoir and if this supply were found to be
of an acceptable purity no further distillation would be required.
This approach was soon found to be completely inadequate. The water
was in constant contact with the aluminum flanges of the reservoir
and no means were available to check the water purity once the
reservoir had been filled.
-104-


Figure 10. NH3 #19a, b, and c: Cubic Phase
NOTES
Frost grown by deposition at high flow rate in order to form cubic phas
NH3 #19a, b, and c are all the same frost which was held at 77 K and
scans were taken at one-hour intervals.
For this frost the flow rate was initially low and when flow was in
creased the transition to cubic phase was very obvious by visual
inspection.


Appendix 5
Calibration of PM Tubes
The total errors in the refelctivity measurements are due to the
uncertainties in the cross-calibration of the PM tubes used to measure
the incident and reflected radiation. No two photoelectric detectors
will generate the same electrical output for a given incident radiation.
Each detector will have a spectral sensitivity characteristic of its
photocathode and electron amplification mechanism.
As shown in Figure 1, the two PM tubes were mounted inside the
frost chamber: one to monitor the incident radiation (PM 9553) and
one to monitor the reflected radiation (PM 6157). There were no
optical components between PM 9553 and the frost and likewise for
PM 6157.
To measure the frost reflectivity it is only necessary to know
the relative sensitivity of the PM tubes since, once the incident
radiation (IQ) is measured by PM 9553, this I0 can be adjusted to what
PM 6157 would have measured. A secondary standard (BaSO^) was employed
to determine the hemispherical reflectivity; therefore, it is not
required that the PM tubes be calibrated on an absolute base. The
relative sensitivity of the PM tubes or cross-calibration must be ex
perimentally determined even though the manufacturer often supplies
the quantum efficiency of each PM tube.
-108-


Figure 16. 1^0 #7a and b: Cubic Phase
NOTES
This frost was grown with a 10 ^ Hg buffer gas (N2).
Both traces are cubic 1^0. Trace "b" is after annealing for one and one-half
hours at T <\,225 K. Note the increased difference in reflectivity of trace
a and b with annealing time. In 1^0 #5 the annealing was done for about one-
half hour.


Figure 17. HO #10: (a), Amorphous Phase; (b), Cubic Phase
NOTES
a) Trace (a) is very fine-grained amorphous ^0.
b) Trace (b) is cubic frost annealed from amorphous frost.
c) The amorphous frost in this experiment may not have been optically thick. The
frost was grown to examine the absorption features at 2200A, 2075A, and 1925&.
d) Chamber was closed.
e) No buffer gas.
f) Amorphous growth period was 1.5 hours.
, -. 6 o
g) The X law is shown by triangles for X > 2200A.


-11-




Figure 19. 1^0 #13A: Amorphous Phase
NOTES
a) This frost grown to examine absorption features in the cubic phase when
NH^ is added on top of H^O cubic.
b) The amorphous frost was grown and reflectivity measured to see if cubic
absorption feature would be present. Only the 1925 dip was seen.
c) Frost grown at very slow flow rate.
d) Growth period 3.5 hours.


-26-
The most striking result was the reflectivity exceeding 100 per
cent for frosts optically thick at 3000&. There are two possible causes
for this result; one is a consistent error in photometry due to PM
calibration errors and the second is that the frosts were not Lambert
reflectors for radiation in the middle ultraviolet. It must also be
mentioned that the different frost thicknesses for each experiment
can introduce a maximum uncertainty of 5 percent for all wavelengths.
For the thicker frosts the PM tube monitoring the reflected light was
closer to the frosts and thus had a larger collecting solid angle
while the PM tube measuring the incident flux always collected the
total light on the frost. This effect can be seen at 3000^ where the
reflectivity differs from frost to frost.
As for the spectral variations in reflectivity the uncertainty
lies in the Lambert assumption or calibration errors. On review of
the cross-calibration curve it is clear that at best the spectral
variance of reflectivity could be flat or increasing toward shorter
wavelengths, corresponding to negative or positive calibration errors,
respectively. As discussed in Appendix 5, the errors in cross
calibration were found to stem from nonuniformities of the PM tube
photocathodes and little could be done to correct this problem; how
ever, it seems safe to conclude that the reflectivities of both NH-j
and 1^0 cubic frosts increased toward shorter wavelengths. Supporting
evidence for this conclusion is the fact that some of the amorphous
frosts were indeed found to be constant in spectral reflectivity in the
wavelength region of no absorption.


Figure 30. Comparison of Jovian UV Albedo to NH^
Frost Reflectivity.


-86-
I


130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800
2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
Figure 25. BaSO^ and Stainless Steel Substrate Reflectivities
i
00
ro
I


-54-
Thus, the result of a buffer gas is to mix the amorphous and cubic
phases and in most instances grow a cubic 1^0 frost overlying an amor
phous H2O. From Figures 15 and 16 it is also clear that since anneal
ing of the buffer-gas-frosts increases the grain size, the shape of
the absorption feature at 1925& is related to the grain size of the
material. No means were available to determine an average grain size
during the experiments so at best it can be concluded that an increased
grain dimension also increases the strength and width of the absorption
features. This grain size effect was also observed by Kieffer (1968)
in the infrared region for CO2 and H2O frosts.
For H2O, #10, 12, 13, and 14 the six-inch diffusion pump was open
to the frost chamber during the growth period so that the amorphous
H2O was easily formed since all noncondensible (buffer) gases are
pumped from the chamber. The amorphous H2O appeared grey and very
fine-grained. It was difficult to obtain an optically thick amorphous
frost since this phase of solid H2O usually grew as a translucent ice.
o
The reflectivity was checked intermittently at 3000A during the growth
period and when it converged toward the BaSO^ reflectivity the growth
was stopped. Typical growth periods ranged from three to five hours.
The annealing procedure outlined previously was employed to
change the amorphous phase to the cubic phase. The H2O phase change
o
occurred at 150 K and was accompanied by a rise in chamber pressure
(see Figure 29). A temperature rise, of the substrate, at the phase
change requires an exothermic reaction within the frost. It thus
follows that the energy level of the amorphous H2O must be greater,
for a given temperature, than the energy level of the cubic H2O for
a release of heat to occur during the phase change. This further


Figure 20. H2O #13: (i), (c), Cubic Phase; (d) NH3 Added Over H2O Cubic
NOTES
a) Traces (b) and (c) are cubic H^O. Trace (c) was annealed a second time. Trace
(d) is reflectivity of 1^0 cubic with a NH^ concentration of approximately .003
mole/liter. The NH^ was added on top of 1^0 cubic.
b) Each annealing of frost (b) and (c) was about 45 minutes.
c) Note the increase in reflectivity when NH^ was added. This could be explained
by an increase in scattering by the fine-grained NH^ particles formed over
coarse 1^0 cubic grains.


106-
To determine the concentration of NH
conductivity is first defined, i.e.,
the equivalent
A K/C
-1 2 -1
where A is the equivalent conductivity [ohm cm equivalents ], K
is the specific conductivity [ohm ^ cm and C is the concentration
_3
[equivalents cm ]. The concentration is measured in equivalents,
where one equivalent is the weight of substance necessary to give one
mole of H or OH in a neutralization reaction. Put another way, the
equivalence is the formula weight divided by the valence. For example,
if the ions that neutralize a one molar solution of H+ have a valence
of two then only .5 moles need be added. If thevvalence is unity the
concentration is in equivalents per liter or simply moles per liter.
For large concentrations of NH^OH the equivalent conductivity
is listed in the Handbook of Chemistry and Physics. The values are:
A r~c
238.00 0.0000
9.66 0.1000
5.66 0.1732
3.10 0.3160
Since NH^OH is a weak electrolyte the equivalent conductivity
increases very rapidly for dilute mixtures. To calculate C a conser
vative estimate of 100 for A is taken and the specific conductivity
was experimentally measured to be .5 x 10~6 ohm-1 cm-1. Thus C = K/A =
-9
5. x 10 equivalents per liter or moles per liter. For NH^OH there
are 35 grams per mole so C = .2 x 10~6 gr/cc.
For a concentration of .2 ppm of NH^ it is highly unlikely that
any feature in the H20 frosts could be caused by NH3>


-13-
The frost dewar was connected to LN£ vacuum feedthroughs to prevent
O-ring freeze out and was fed from a 25 liter LN^ supply dewar. The
temperature of the frost dewar was monitored by an iron-constantan
thermocouple silver soldered to its front surface.
D. Photometry
Two EMR 541F-05M-18 solar-blind photomultiplier tubes (hereafter
denoted PM tubes) were positioned within the frost chamber to record
the incident and reflected UV radiation. These PM tubes are sensitive
to radiation with wavelengths between 340oX and 140oX. Each PM tube
output was connected to a Fairchild solid-state preamplifier with a
xl gain and a low pass filter. After amplification and filtering,
the PM tube output was displayed on a chart recorder and a digital
volt meter. High voltage was provided by a Fluke 0-6000 volt power
supply.
One of the PM tubes could be moved remotely into the beam to
record the total incoming flux while the second PM tube was mounted
to collect the reflected light at approximately 10 from normal in
cidence (see Figure 1). The monochromator was dialed to the desired
wavelength and the total incident flux was measured. The PM tube
used to record this signal was then moved out of the beam and the
reflected light measured. This basic procedure was continued until
all wavelengths were covered. Wavelength steps of 100& were used if
the frost reflectivities were a continuum (X > 2400^) and steps of
o
25A were used if reflectivities exhibited features and absorption
cutoffs.


MOLAR ABSORPTION COEFFICIENT (LITER MOLE CM
-46-
2200
2000 1800 1600
WAVELENGTH (l )
1400
Figure 12. NH^ Solid Absorption Coefficients (from
Dressier and Schnepp, 1960)


REFLECTIVITY (%)
1
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
&
*


IV. Conclusions
The reflectivities of NH3 and H2O frosts were measured from
o o
3000A to 1400A and the effects on reflectivity of grain growth,
cubic or amorphous phase, and buffer gas were examined.
The reflectivites of the NH3 frosts are above 90 percent long-
o 00
ward of 2300A and below 1 percent from 1950A to 1400A. The absorp-
o o
tion cutoff from 2200A to 1950A occurs at longer wavelengths than
those expected from the absorption coefficients of solid NH3 measured
by Dressier and Schnepp (1960).
The H2O frost reflectivity measurements are quite different
from anything anticipated. Based on absorption data of Dressier and
Schnepp, a H2O frost should not become "black" until approximately
o
1500A. The present study shows that if the H2O frosts are optically
thick at 300oX the reflectivity will decrease to less than 10 percent
o o
between 1800A and 1700A. This aspect is more important in relation
to the study of the Jovian planets, in particular Saturn's rings.
o
Since the solar radiation at 1800A is approximately an order of magnitude
o
greater than at 1500A, detection of a solid H2O absorption will be
easier than previously thought.
It was also discovered that the H2O cubic frosts contained
absorption features which are not seen in 1^0 vapor. The possibility
-90-


ACKNOWLEDGMENTS
The author wishes to express his gratitude to Dr. R. C. Anderson,
whose technical assistance, professional interest, and efforts put
forth in obtaining the financial support for this study were far be
yond the requirements of a committee chairman.
The author would like to thank Dr. A. E. S. Green, Dr. A. G.
Smith, Dr. B. M. Leadon, Dr. M. H. Clarkson, and Dr. D. T. Williams
for their efforts contributed as members of his supervisory committee.
Special acknowledgment is extended to Dr. T. G. McRae for his
complete and accurate technical advice on vacuum system techniques.
The author would also like to acknowledge Mr. H. E. Stroud for
his general assistance in procuring equipment and materials needed
for the construction of the experimental apparatus.
This research was supported by the National Science Foundation,
Grant GA28852.
iii


-114-
Wallace, L., J. J. Coldwell, and B. D. Savage, 1972: Ultraviolet
Photometry from the 0A0-1II Observations of Venus, Mars,
Jupiter, and Saturn Longward of 2000X, The Astrophysical J.,
172.
Watanabe, K., M. Zelikoff, and E. C. Y. Inn, 1953: AFCRC Tech. Rpt.
No. 53-23, Geophys. Res. Paper No. 21.
Wood, B. E. and A. M. Smith, 1968: Spectral Reflectance of Water and
Carbon Dioxide Cryodeposits from .36 to 1.15p, AIAA J. T_, 1362-
1367.
Wood, B. E., A. M. Smith, J. A. Roux, and B. A. Seiber, 1971:
Spectral Infrared Reflectance of HoO Condensed on LN2~Cooled
Surfaces in Vacuum, AIAA J., 9^, 1836-1842.


To my wife, Betty


BIOGRAPHICAL SKETCH
John G. Pipes was bom on September 6, 1945, in Gardner, Massa
chusetts. He graduated from Dan McCarty High School, Ft. Pierce,
Florida, in June, 1963. In September, 1963, he entered Indian River
Junior College in Ft. Pierce and received the Associate of Arts degree
two years later. In September, 1965, he transferred to the University
of Florida, where he received the degree of Bachelor of Science in
Aerospace Engineering in December, 1967. In January, 1968, he enrolled
in the Graduate School of the University of Florida. He spent April
and May of 1968 as a visiting graduate student at Kitt Peak National
Observatory in Tuscon, Arizona during the preparation of an Aerobee
rocket package for the observation of the ultraviolet spectrum of
Jupiter.
In June of 1969 he received the Master of Science in aerospace
engineering, having done graduate work in planetary atmospheres and
engineering optics. From that date until the present he has pursued
the degree of Doctor of Philosophy at the University of Florida.
John G. Pipes is married to the former Betty Anne Veber and
has two children, Michael, age four, and Pamela, age two.
-115-


-103-
conduction heat load from the chamber walls and thus the frost was
exposed to "wall temperatures" of 100 K instead of 300 K. The shield
also served to keep any frost fragments from falling into the vacuum
pump and to attach several additional components, e.g., the BaSO^
reference and an observation light.
Components of the vacuum system employed to pump the frost
chamber were the following:
a) Six-inch oil diffusion pumps manufactured by
Norton Vacuum Equipment Division. The oil
used was DC-705.
b) Six-inch Chevron Cryo-Baffle was the cold trap
for the diffusion pump.
c) An air-operated 6" gate valve separated the
diffusion pump and cold trap from the frost
chamber. This gate valve was controlled by
the power failure safety-shutoff system which
also controlled the gate valve on the mono
chromator vacuum system.
d) The diffusion pump was backed by a 15 cfm
Duo-Seal mechanical vacuum pump. This same
pump was connected via vacuum valves to the
diffusion pump for the McPherson monochromator.
The vacuum grease used on all parts was Apiezon L.


130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600
WAVELENGTH (A)
2 8 00
3 000
I
Ul
vO
!


REFLECTIVITY {%)
J
130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
-89-


-27-
It is just as possible that the frosts do not diffusely reflect
as a Lambert surface for UV wavelengths and have a scattering phase
function somewhat characteristic of a Rayleigh scattering media. This
would result in a higher reflectivity near normal incidence than that
of a Lambert reflector (e.g., the BaSO^ reference). Unfortunately this
possibility of an unknown scattering phase function could not be ex
amined in this experiment since it would require recording the reflectiv
ity at all reflecting angles, a task beyond the capabilities of the
system.
For the wavelength region below 2400X the ammonia frosts were
easily divisible into three groups: Group 1: NH3 #9, 10, and 19 are
cubic frosts that were grown directly into the cubic phase, Group 2:
NH3 #11 and 17b are cubic frosts formed by warming the amorphous
deposits until they were within the phase change temperature range
(when data were taken the frosts were recooled to 77 K), and
Group 3: NH^ #12 and 17a are amorphous frosts.
Group 3 shows no structure other than a continuum type cutoff
from 2400& to 1950$ and then was black out to 1400& (at I6O0X the
light source had sufficient output so that some return light could be
measured and the result was a reflectivity less than 1 percent which
was termed "black").
The Groups 1 and 2 were both cubic NH3 frosts but Group 2
showed a reduction of 20 percent in reflectivity between 2400& and
o
2200A and then a decrease of only 10 percent in reflectivity for the
next lOoX prior to the sharp cutoff between 2100& and 2000X. Why this


-113-
Kief fer H. H., 1968: Near Infrared Spectral Reflectances of Simu
lated Martian Frosts, Dissertation, California Institute of
Technology.
, 1969: Reflectance Spectrometer/Environmental Chamber
for Frosts, Applied Optics, 2 2497-2500.
, 1970: Spectral Reflectance of CO2-H2O Frosts, J. of
Geophysical Research, 75, 501-9.
Knox, R. S., 1963: Theory of Excitons, Academic Press (Supplement 5).
Kuiper, G. P., D. P. Cruikshank, and U. Fink, 1970a: The Composition
of Saturn's Rings, Sky and Telescope, 39^, 14.
, 1970b: (letter to editor), Sky and Telescope, 22. 80.
Lewis, J. S., 1969: The Clouds of Jupiter and the NH3 -H2O and NH3 -H2S
System, Icarus, J^ 365-378.
Mauer, F. A., L. H. Bolz, H. S. Peiser, and H. F. McMurdie, 1972:
(private communication, Notes on non-cubic NH^).
Owen, T., 1965: Saturn's Ring and the Satellites of Jupiter: Inter
pretations of Infrared Spectra, Science, 149, 974-5.
Pilcher, C. G., C. R. Chapman, L. A. Lebofsky, and H. H. Kieffer,
1970: Saturn's Rings: Identification of Water Frost, Science,
167, 1372-3.
Prinz, G. A., 1972: (private communications), U. S. Naval Research
Laboratory.
Schnepp, 0. and K. Dressier, 1960: Absorption Spectra of Solid Xe, Hr,
and Ar in the Vacuum Ultraviolet, J. of Chemical Physics, 33,
49-55.
Seiber, B. A., B. E. Wood, A. M. Smith, P. R. Muller, 1970: Density of
Low Temperature Ice, Science, 170, 652-4.
Smith, A. M., K. E. Tempelmeyer, P. R. Muller, and B. E. Woods, 1969:
Angular Distribution of Visible and Near I. R. Radiation Re
flected from CO2 Cryodeposits, AIAA Journal, 7_, 2274-80.
Van De Hulst, H. C., 1957: Light Scattering by Small Particles, John
Wiley & Sons, Inc.


Figure 28. "Ball" Frost Growth Sequence
NOTES
These photographs were taken at 30-minute intervals during the growth
period. The buffer gas is ^ at 500 jj. Note in photograph "A" the
underlying amorphous 1^0 formed at 77 K during start of flow.


REFLECTIVITY (%)
1400 1600 1800 2000 2200 2400 2600
WAVELENGTH (A)
2600
3000
t


III. H2O and NH3 Frost UV Reflectivities
A. Frost Growth Procedures
The techniques of controlling the growth environment for both
NH3 and H2O were strictly a result of trial and error. The understand
ing of how to control effectively the closely coupled parameters of
latent heat, heat transfer characteristics of the frosts, vapor pres
sure-temperature relationships, and phases of the solids was soon found
to be more difficult than the recording of photometric data. It be
came increasingly obvious that the initial growing conditions (flow
rate, chamber pressure, substrate roughness) dictated to a large ex
tent the growth patterns for the remaining growth period. It was also
recognized that it would be difficult to define these initial condi
tions. Particularly for NH3, the establishment of whether the frost
was in an amorphous or a cubic phase, or a combination of these two,
was a major experimental problem.
From the works of Dressier and Schnepp (1960), Mauer et al.
(1972), and Black et al. (1958), the techniques for obtaining an
essentially complete amorphous phase are well established. The data
of Dressier and Schnepp show that the solid cubic ammonia begins to
o
absorb about 200A deeper into the UV than the amorphous solid (see
Figure 12). Early data taken herein always showed that both the
amorphous and cubic frosts absorb strongly between 220oX and 2000X.
-23-


-48-


130
120
110
100
90
80
70
60
SO
40
30
20
10
0
1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)
VO
LJ
I


I
BIBLIOGRAPHY
Anderson, R. C., J. G. Pipes, A. L. Broadfoot, and L0 Wallace, 1969:
Spectra of Venus and Jupiter from 1800A to 3200A J. of Atm. Sci.,
26, 874-888.
Anderson, R. C. and J. G. Pipes, 1971: Jovian Ultraviolet Reflectivity
Compared to Absorption by Solid Ammonia, J. of Atm. Sci., 28,
1086-7.
Benford, F., S. Schwarz, and G. P. Lloyd, 1948a: Coefficients of Re
flection in the Ultraviolet of Magnesium Carbonate and Oxide,
J. of the Optical Society of America, 38, 964-5.
Benford, F., G. P. Lloyd, and S. Schwarz, 1948b: Coefficients of Re
flection of Magnesium Oxide and Magnesium Carbonate, J. of the
Optical Society of America, 35, 445-447.
Billmeyer, F. W., Jr., 1969: Part X: White Reflectance Standards,
Optical Spectra, Jan/Feb.
Black, I. A., L. H. Bolz, F. P. Brooks, F. A. Mauer, and H. S. Peiser,
1958: A Liquid-Helium Cold Cell for Use with an X-ray Diffracto
meter, J. of Research of the National Bureau of Standards, 61,
367-371.
Dressier, K. and 0. Schnepp, 1960: Absorption Spectra of Solid Methone,
Ammonia and Ice in the Vacuum Ultraviolet, J. of Chemical Physics,
33, 270-274.
Eastman Kodak, 1969: 'Eastman White Reflectance Paint,' Eastman White
Reflectance Standard, #JJ-32 (paint) and JJ-31 (standard).
Fehsenfeld, F. C., K. M. Evenson, and H. P. Broida, 1964: NBS Report
#8701, Microwave Discharge Cavities Operating at 2450 MHZ.
Grum, F. and G. W. Luckey, 1968: Optical Sphere Paint and a Working
Standard of Reflectance, Applied Optics, _7, 2289-94.
Harrison, H. and R. I. Schoen, 1967: Evaporation of Ice in Space:
Saturn's Rings, Science, 157, 1175-6.
.-112-


130
120
110
100
90
80
70
60
50
40
30
20
10
0
1400 1600 1800 2000 2200 2400 2600 2800 3000
WAVELENGTH (A)