From: "St-Onge, Louis" <louis.st-onge@nrc.ca>
Newsgroups: sci.physics.plasma
Subject: Re: H alpha line profile
Date: Thu, 05 Feb 1998 09:43:39 -0500
Organization: CNRC/IMI
Message-ID: <34D9D01B.322@nrc.ca>
References: <34D86CEC.1F64@sci.muni.cz>
Reply-To: louis.st-onge@nrc.ca


Antonin Brablec wrote:
>
> Dear colleagues,
> we investigate a special type of corona discharge burning in liquid
> water
> and we try to estimate electron density and temperature from line shape
> of
> H alpha. We try to use the standard Griem's tables (impact broadening by
> electrons and static broadening by ions) for fitting with the measured
> line
> profile in order to verify that such procedure can be also used in this
> case.
> However, the measured profile of H alpha is a little bit asymmetrical
> and
> so the fitting is not good (especially on wings). It could indicate that
> other additional broadening effects can play important role.
> So, we are interested in other broadenings like are resonant, van der
> Waals,
> etc. and in effect which can produce asymmetrical line shape. Do you
> have
> experience with the matter? We also looking for analytical formula of
> the mentioned broadenings (or tabulated values or programs).
> Can somebody advice us?

Dear colleague,

The H alpha line is intrinsically asymmetrical. This is due to the fine
structure of the levels involved in the transition, not to a broadening
mechanism. In fact, H alpha is composed of seven separate transitions,
given below in L-S coupling notation, with the wavelength positions
(relative to the most intense transition), and the relative intensities:

Transition Wavelength shift (nm) Normalized intensity

3d 2D3/2 --- 2p 2P1/2 -0.01418 0.556
3d 2D3/2 --- 2p 2P3/2 0.00155 0.111
3d 2D5/2 --- 2p 2P3/2 0.0 1.0
3p 2P1/2 --- 2s 2S1/2 -0.00802 0.116
3p 2P3/2 --- 2s 2S1/2 -0.01267 0.231
3s 2S1/2 --- 2p 2P1/2 -0.00996 0.0109
3s 2S1/2 --- 2p 2P3/2 0.00577 0.0217

In the past, I did some work on microwave hydrogen discharges (in the
1-10 torr range) and determined the gas temperature by measuring the
Doppler broadening of H alpha (using a Fabry-Perot interferometer). I
had to take into account the fine structure of the line because
otherwise I would have overestimated the temperature. Indeed, not only
does the fine structure cause an asymmetry of the line but also a
broadening (because of the wavelenght shifts between the different
components). For example, for a temperature of about 630 K, each
component of the line has a width of 0.275 cm-1, but the resulting line
(obtained by adding up the seven components) has a width of 0.55 cm-1.
In this case, the overestimation would be by a factor of two. Of course,
the larger the broadening of each component, the smaller is the
influence of the fine structure on the measured width.

I don't know about broadening mechanisms in corona discharges. You will
have to determine how important is the influence of the fine structure
of H alpha in your case.

Finally, a reference I found useful at the time was:

K Matsunaga, K Kadota, M Fujiwara, and J Fujita, Jpn. J. Appl. Phys. 20
(1981) L615.

Hope this helps,

Louis St-Onge, Ph.D.
Process Instrumentation
Industrial Materials Institute
National Research Council Canada
75 de Mortagne Blvd.
Boucherville, Québec J4B 6Y4, Canada

E-mail: louis.st-onge@nrc.ca