The Holstein‐Biberman theory of the transport of resonance radiation in gases is generalized to include the case in which the boundary is partially reflecting. The problem is formulated in a manner which follows conventional transport theory somewhat more closely than the earlier treatments of Holstein and Biberman. Using the incoherent scattering approximation, we first write down the two coupled Boltzmann equations describing the mutual transport of excited atoms and resonance photons. The photon equation is then solved, with (diffuse) reflecting boundary conditions, by a slight extension of the interreflection method, and inserting the result into the excited atom equation leads at once to the appropriate generalization of the Holstein‐Biberman transport equation. The kernel of this integrodifferential equation, G(r,r′), represents the mean probability that a resonance quantum emitted at r′ is absorbed at r, taking into account the contribution of paths which strike the boundary an arbitrary number of times; G(r,r′) is expressed quite generally in terms of the resolvent kernel of the interreflection equation. The theory is used to calculate the effect of a slightly reflecting boundary on the imprisonment time in a gas discharge.

1.
K. T.
Compton
,
Phys. Rev.
20
,
283
(
1922
);
E. A.
Milne
,
J. London Math. Soc.
1
,
1
(
1926
);
C.
Kenty
,
Phys. Rev.
42
,
823
(
1932
);
S. Chandrasekhar, Radiative Transfer (Clarendon Press, Oxford, 1950);
F. N.
Edmonds
,Jr.
,
Phys. Rev.
78
,
424
(
1950
);
a discussion of averaging procedures and extensive references to the earlier literature on the subject are given in A. C. G. Mitchell and M. W. Zemansky, Resonance Radiation and Excited Atoms (The Macmillan Company, New York, 1934).
2.
For a discussion of the astrophysical applications and for references to the astrophysical literature, see, for example, V. A. Ambartsumyan, Theoretical Astrophysics (Pergamon Press, New York, 1958).
3.
P. F. Little, Proceedings of the Fourth International Conference on Ionization Phenomena in Gases (North‐Holland Publishing Company, Amsterdam, 1960).
4.
L. M.
Biberman
,
J. Exptl. Theoret. Phys. (U.S.S.R.)
17
,
416
(
1946
).
5.
T.
Holstein
,
Phys. Rev.
72
,
1212
(
1947
);
T.
Holstein
,
83
,
1159
(
1951
).,
Phys. Rev.
6.
L. M.
Biberman
and
I. M.
Gourevitch
,
J. Exptl. Theoret. Phys. (U.S.S.R.)
19
,
507
(
1949
).
7.
D.
Alpert
,
A. O.
McCoubrey
, and
T.
Holstein
,
Phys. Rev.
76
,
1257
(
1949
);
D.
Alpert
,
A. O.
McCoubrey
, and
T.
Holstein
,
85
,
985
(
1952
).
8.
R.
Siewert
,
Ann. Physik
17
,
371
(
1956
).
9.
P. J.
Walsh
,
Phys. Rev.
107
,
338
(
1957
).
10.
A. V.
Phelps
,
Phys. Rev.
110
,
1362
(
1958
).
11.
A. V.
Phelps
and
A. O.
McCoubrey
,
Phys. Rev.
118
,
1561
(
1960
).
12.
Notably, the fluorescent lamp discharge. Recently,
K. H.
Butler
and
H. H.
Homer
,
Illum. Eng.
55
,
396
(
1960
), have given experimental evidence which they interpret, in terms of a crude but not unreasonable model, as meaning that as far as the ultimate production of useful resonance radiation is concerned, a surprisingly large fraction of the energy reflected back into the discharge by the phosphor coating on the wall is wasted; further, they suggest that it is mainly this effect which determines the optimum particle size distribution of the phosphor coating. Without a theory which takes into account the effect of the reflecting wall on the transport process (and hence on the characteristics of the discharge itself) it is very difficult to assess the correctness of this interpretation.
13.
That is, the average time required for a resonance quantum to escape from the enclosure (see reference 9). In the fluorescent lamp application, the imprisonment time of the Hg λ2537 resonance radiation is one of the most important parameters governing the behavior of the discharge. See, for example,
C.
Kenty
,
J. Appl. Phys.
21
,
1309
(
1950
);
J. F.
Waymouth
and
F.
Bitter
,
J. Appl. Phys.
27
,
122
(
1956
);
M. A.
Cayless
,
Brit. J. Appl. Phys.
10
,
186
(
1959
).
14.
In what follows, we assume k(ν) to be constant over the enclosure. Only formal modifications are required if, because of a temperature gradient, for example, k(ν) varies with position.
15.
V. V. Sobolev, Vestnik Leningrad. Univ. No. 5, 85 (1955) (AEC‐tr‐2738).
16.
With the above approximations, the behavior of n(r) at the boundary of the enclosure is determined by the boundary conditions we impose on the photon distribution.
17.
P. B.
Moon
,
J. Opt. Soc. Am.
30
,
195
(
1940
).
18.
M. A.
Weinstein
,
J. Opt. Soc. Am.
51
,
723
(
1961
).
19.
In what follows, we assume the surface defining the boundary of the enclosure to be nonreentrant, and we exclude consideration of intrinsic singular points which may require separate statements about isolated points or lines.
20.
See, for example, F. G. Tricomi, Integral Equations (Interscience Publishers, Inc., New York, 1957).
21.
Since (dn/dt)i depends only on internal processes, it must have the symmetry of the enclosure.
22.
See, for example, A Erdélyi, W. Magnus, F. Oberhettinger, and F. G. Tricomi, Higher Transcendental Functions (McGraw‐Hill Book Company, Inc., New York, 1953), Vol. 2, p. 5, Eqs. (12) and (13).
23.
If there are metastable levels near the resonance level, Eq. (69) is still correct as long as one can neglect the diffusion of metastables compared with the transitions among the metastable, resonance and ground states; ν′ and ν are then algebraic functions of the average cross sections for these transitions (see reference 9).
24.
We are neglecting the temperature gradients in the gas.
25.
The approximation for ΔTi/Ti is considerably better than for Ti itself.
This content is only available via PDF.
You do not currently have access to this content.