Interstellar Gamma Ray Lines from Low Energy Cosmic Ray Interactions
Extracted source text ยท arXiv astro-ph/9606034
\documentstyle{l-aa}
\def\lsim{\lower.5ex\hbox{$\; \buildrel < \over \sim \;$}}
\def\gsim{\lower.5ex\hbox{$\; \buildrel > \over \sim \;$}}
\begin{document}
\thesaurus{13 (13.07.2; 09.03.2)}
\title{Interstellar Gamma Ray Lines from Low Energy Cosmic Ray Interactions}
\author{Reuven Ramaty
%\inst{1}
}
%
\institute{Laboratory for High Energy Astrophysics,
NASA Goddard Space Flight Center,
Greenbelt, MD 20771, USA
}
\date{Received ; accepted }
\offprints{R. Ramaty}
\maketitle
\markboth{Ramaty: Interstellar Gamma Ray Lines from Low Energy
Cosmic Ray Interactions}{}
\begin{abstract}
Evidence for the existence of low energy cosmic rays in the Galaxy
comes from the COMPTEL observations of gamma ray line emission from
Orion, and also from light element abundance data which seem to
suggest a low energy rather than a relativistic Galactic cosmic ray
origin for most of the light elements. The Orion and light isotope
data are more consistent with a composition that is depleted in
protons and $\alpha$ particles than with one which is similar to that
of the Galactic cosmic rays. This low energy cosmic ray phenomenon
appears to be highly localized in space and time in the Galaxy, and
is probably associated with star formation regions similar to Orion.
\keywords{Gamma Rays: theory -- Galaxy: abundances -- Nuclear
reactions, nucleosynthesis, abundances}
\end{abstract}
\section{Introduction}
The interactions of accelerated particles with ambient matter
produce a variety of gamma ray lines following the deexcitation
of excited nuclei in both the ambient matter and the
accelerated particles. Apart from solar flares, nuclear
deexcitation lines following accelerated particle interactions
have so far only been observed from the Orion molecular cloud
complex (Bloemen et al. 1994). The Orion observations, carried
out with the COMPTEL instrument on the Compton Gamma Ray
Observatory (CGRO) revealed emission features at 4.44 and 6.13
MeV, due to deexcitations in $^{12}$C and $^{16}$O. Since there
are no significant long lived radioactive isotopes of
nucleosynthetic origin that decay into the excited states of
these nuclei, the observed lines must be produced
contemporaneously by large fluxes of accelerated particles
interacting in Orion. Gamma ray emission in the energy range
from about 30 MeV to 10 GeV was also observed from Orion with
the EGRET instrument on CGRO (Digel, Hunter, \& Mukherjee
1995). However, the fact that these EGRET data do not require
that the cosmic rays in Orion be enhanced relative to the
relativistic Galactic cosmic rays observed near Earth, implies
that the particles which produce the line emission in Orion are
mostly low energy cosmic rays confined to energies below the
effective pion production threshold.
The discovery of gamma ray line emission from Orion and the
implied existence of large fluxes of low energy cosmic rays,
not only in Orion but probably also elsewhere in the Galaxy,
has led to renewed discussions on the origin of the light
elements. It has been known for over two decades that cosmic
ray spallation is important to the origin of $^6$Li, $^7$Li,
$^9$Be, $^{10}$B, and $^{11}$B (Reeves, Fowler, \& Hoyle 1970).
It was shown that the relativistic Galactic cosmic rays (GCR)
interacting with interstellar matter prior to the formation of
the solar system may have produced the observed solar system
abundances of $^6$Li, $^9$Be and $^{10}$B (Meneguzzi, Audouze
\& Reeves 1971; Mitler 1972). These cosmic rays, however,
cannot account for the abundances of $^7$Li and $^{11}$B. It is
believed that about 10\% of the $^7$Li is produced in the big
bang and that most of the remaining production is due to
nucleosynthesis in stars (e.g. Reeves 1994). Recent
measurements of the boron isotopic ratio, $^{11}$B/$^{10}$B, in
meteorites yielded values in the range 3.84 -- 4.25 (Chaussidon
\& Robert 1995) which exceed the calculated GCR value by a
factor of about 1.5. The implications of the Orion gamma ray
observations on the origin of the light isotopes have been
considered by Cass\'e, Lehoucq \& Vangioni-Flam (1995) and
Ramaty, Kozlovsky \& Lingenfelter (1996). Unlike the GCR, low
energy cosmic rays, similar to those which produce gamma ray
line emission in Orion, could account for the meteoritic B
isotopic ratio. It is in fact possible that the bulk of the
light isotopes, except $^7$Li, is produced by low energy cosmic
rays in the Galaxy. Although $^{11}$B could have been produced
by neutrino spallation of $^{12}$C in supernovae (Woosley et
al. 1990), recent B and Be observations in stars of various
metallicities (Duncan 1995) support a low energy cosmic ray
origin for $^{11}$B since neutrino spallation is not expected
to produce much Be.
The present paper is in large part a review based on a series
of previous articles dealing with the gamma ray line emission
from Orion and related subjects (Ramaty 1995; Ramaty, Kozlovsky,
\& Lingenfelter 1995a,b;1996). After a general discussion, we
consider the implications of the gamma ray line observations on
the composition, energy spectrum, energy deposition and energy
density of the low energy cosmic rays in Orion. We also
consider the ionization produced by these cosmic rays, and
briefly review the possible origins of the accelerated
particles. We then proceed to calculate the expected gamma ray
line emission produced by low energy cosmic rays in the Galaxy.
We base this calculation on the close relationship between the
gamma ray line and light isotope production.
\section{General Considerations}
The gamma ray lines produced by accelerated particle
interactions can be broad, narrow or very narrow (Ramaty,
Kozlovsky, \& Lingenfelter 1979). Broad lines are produced by
accelerated C and heavier nuclei interacting with ambient H and
He. The broadening of these lines (widths ranging from a few
hundreds of keV to an MeV) is due to the motion of the
accelerated heavy particles themselves. Narrow lines are
produced by accelerated protons and $\alpha$ particles
interacting with ambient He and heavier nuclei. The broadening
in this case (widths ranging from a few tens of keV to around
100 keV) is due to the motion of the heavy targets which recoil
with velocities much lower than those of the projectiles. Very
narrow lines result from excited nuclei which have slowed down
and stopped due to energy losses before emitting gamma rays.
The broadening of these lines is due only to the bulk motion of
the ambient medium (widths around a few keV or less for the
interstellar medium).
There are two distinct processes which can lead to very narrow
line emission: deexcitation of heavy nuclei embedded in dust
grains and excited by protons or $\alpha$ particles
(Lingenfelter and Ramaty 1976), and deexcitation of excited
nuclei populated by long lived radionuclei. Line emission from
dust is not discussed in the present article. Dust grains,
however, may play an important role in the injection and
acceleration of the low energy cosmic rays which produce the
gamma ray lines (see below). Long lived radionuclei produced by
accelerated particle bombardment can stop in ambient gas before
they decay thereby producing excited nuclei essentially at
rest. The most important such radionuclei are
$^{55}$Co($\tau_{1/2}=17.5$h), $^{52}$Mn($\tau_{1/2}=5.7$d),
$^{7}$Be($\tau_{1/2}=53.3$d), $^{56}$Co($\tau_{1/2}=78.8$d),
$^{54}$Mn($\tau_{1/2}=312$d), $^{22}$Na($\tau_{1/2}=2.6$y), and
$^{26}$Al($\tau_{1/2}=0.72$my), all of which can be produced in
accelerated particle interactions, for example
$^{56}$Fe(p,n)$^{56}$Co. Unlike the very narrow grain lines
which are produced almost exclusively by accelerated protons
and $\alpha$ particles, very narrow lines from long lived
radioactivity can result from both these interactions and
interactions due to accelerated heavy nuclei.
In the following discussion we shall need to make assumptions
on the composition of both the ambient medium and the
accelerated particles, on the energy spectrum of the
accelerated particles, and on the interaction model. As these
have been described in detail in Ramaty et al. (1996), we only
give a brief summary here. For the ambient medium we assume a
solar system composition (Anders \& Grevesse 1989). For the
accelerated particles we consider six different composition:
solar system (SS), cosmic ray source (CRS), the ejecta of
supernovae of 35M$_\odot$ and 60M$_\odot$ progenitors (SN35 and
SN60), the winds of Wolf-Rayet stars of spectral type WC, and
pick-up ions resulting from the breakup of interstellar grains
(GR). The grain case is analogous to the anomalous component of
the cosmic rays observed in interplanetary space (Fisk,
Kozlovsky, \& Ramaty 1974; Adams et al. 1991). Interstellar
neutral atoms that penetrate into the solar cavity are picked
up by the magnetized solar wind after being ionized by solar UV
and charge exchange with solar wind protons. Because in the
frame of the wind the ions acquire considerable energy during
the pick up process ($m_i V^2/2$ where V is the speed of the
wind), they form a seed population that is much more easily
accelerated than the rest of the ambient plasma. For Orion it
is conceivable that the equivalent incoming matter is
essentially neutral dust that is broken up by evaporation,
sputtering or other processes. The assumed GR abundances are SS
abundances modified by depletion factors (Sofia, Cardelli \&
Savage 1994). The noble gas (He, Ne, Ar) abundances are set to
zero and H/O = 2, assuming that the bulk of the H and O are in
ice. The acceleration of pick up ions in Orion was also
considered by Ip (1995).
The essential properties of these compositions are the
following: SS, CRS and to some extent SN35 have large proton
and $\alpha$ particle abundances relative to C and heavier
nuclei. On the other hand, because of prior mass loss for SN60
and WC, and because H and He are essentially absent in dust,
the SN60, WC and GR compositions are very poor in protons and
$\alpha$ particles. There is no Ne in the GR composition. The
WC composition is dominated by C and O and has some additional
amount of $^{22}$Ne.
All the calculations are carried out in a thick target model in
which accelerated particles with given energy spectrum and
composition are injected into an interaction region where they
produce nuclear reactions as they slow down due to Coulomb
interactions to energies below the thresholds of the various
reactions. Energetically this is the most efficient way of
producing the nuclear reactions; if the particles are allowed
to escape at higher energies, then energy that would otherwise
be available for producing nuclear reactions, is removed from
the system rendering the process less efficient. Energetic
efficiency is important for gamma ray line production in Orion
because even under optimal conditions the deposited energy, and
the accompanying ionization if the medium is neutral, are very
large. A model in which the particles escape from the
interaction region with negligible energy loss is referred to
as a thin target model. In a thick target the energy spectrum
of the interacting particles is flatter (because of the energy
losses) than their source spectrum; in a thin target the
interacting particle spectrum is identical to the source
spectrum. A 'thin target' situation can also arise without
escape, namely when continuous particle acceleration
compensates for the energy losses so that the spectrum of the
interacting particles remains identical to their source
spectrum.
Because the Coulomb energy losses depend on the charge of the
particles, these losses reduce the importance for gamma ray
production of heavy nuclei relative to lighter nuclei. Thus,
for the compositions poor in protons and $\alpha$ particles
(i.e. SN60, WC and GR), for which practically all the gamma ray
production is due to C and heavier nuclei, the ratio of the
line emission from Ne, Mg, Si and Fe to that from C and O is
smaller in a thick target than in a thin target. This has an
observable consequence: because C and O produce lines
predominantly in the 4--7 MeV region while Ne--Fe produce lines
at energies below 3 MeV, for identical source spectra and
compositions, deexcitation line production below 3 MeV relative
to that above this energy is smaller in a thick target than in a
thin target. This is relevant for Orion, where, as we shall
see, emission in the 1--3 MeV was strongly suppressed relative
to the emission between 4--7 MeV.
The calculations that we present are carried out with two
spectral forms for the accelerated particle source function,
%
\begin{eqnarray}
{dN_i \over dt}(E) = K_i \Big({E \over E_0} \Big)
^{-1.5}e^{-E/E_0},
\end{eqnarray}
%
\noindent and
%
\begin{eqnarray}
{dN_i \over dt}(E) = K_i \Big({E \over E_c} \Big)^{-s} ~~{\rm
for}~~ E>E_c \nonumber\\
= K_i ~~{\rm for}~~ E \le E_c,
\end{eqnarray}
%
\noindent where the $K_i$'s are proportional to the accelerated
particle source abundances, $E$ is energy per nucleon, and the
parameters $E_0$ and $E_c$ are allowed to vary over the broad
range from 2 to 100 MeV/nucl. The nonrelativistic spectral
index of 1.5 in Eq.~(1) is the consequence (e.g. Ellison \&
Ramaty 1985) of shock acceleration with a compression ratio at
its maximal value of 4; the exponential turnover characterizes
the effects of a finite acceleration time or a finite shock
size. There is little theoretical basis for the flat spectrum
given by Eq.~(2). It is a simple form that has been used in
previous calculations of gamma ray line and light element
production (e.g. Ramaty et al. 1995a; Cass\'e et al. 1995;
Ramaty et al. 1996), and we use it here as well; as in the
previous studies (Ramaty et al. 1995a; 1996) we take $s$ = 10.
We refer to the spectrum given by Eq.~(1) as the strong shock
spectrum and to that given by Eq. (2) as the flat spectrum.
\begin{figure}
\picplace{5.7 cm}
\caption{Gamma ray observations from Orion; below 30 MeV --
COMPTEL data (Bloemen et al. 1994); above 30 MeV -- EGRET data
(Digel et al. 1995).}
\end{figure}
\section{Gamma Ray Line Emission from Orion}
Gamma ray line emission in the 3 to 7 MeV range was observed
from the Orion complex with COMPTEL (Bloemen et al. 1994). We
show the COMPTEL data in Fig.~1 together with the higher energy
gamma ray emission observed from Orion with EGRET (Digel et al.
1995). The 3--7 MeV observations show emission peaks near 4.44
and 6.13 MeV, consistent with the deexcitations in $^{12}$C and
$^{16}$O following accelerated particle interactions. This
implies that the ambient matter in Orion is undergoing
bombardment by an unexpectedly intense, locally accelerated,
population of low energy cosmic rays. At other photon energies
the COMPTEL observations reveal only upper limits. As we shall
see, the 1--3 MeV upper limit places strong constraints on the
composition of the low energy cosmic rays. Digel et al. (1995)
have shown that the EGRET data can be understood in terms of a
relativistic cosmic ray flux similar to that observed near
Earth producing gamma rays in Orion via pion decay and
bremsstrahlung. This implies that the low energy cosmic rays
should be confined to energies below the effective pion
production threshold (see also Cowsik \& Friedlander 1995). In
the following we shall be concerned with the properties and
effects of these low energy cosmic rays.
\begin{figure}
\picplace{9.5 cm}
%\epsfxsize=6cm
%\epsffile[0 0 100 100]{dot.ps}
\caption{Calculated gamma ray spectra for the cosmic ray source
and the grain compositions, showing a mixed broad-narrow line
spectrum (upper panel) and a broad line spectrum (lower
panel); the very narrow lines in the lower panel are from long
lived radionuclei. The predicted 0.847 MeV line flux in
the upper panel is 5$\times$10$^{-6}$ ph cm$^{-2}$ s$^{-1}$.}
\end{figure}
\subsection{Composition}
In all the calculations of gamma ray production in Orion the
ambient gas was assumed to have a solar system composition
(Anders \& Grevesse 1989). This should not suggest that there
could be no departures from solar abundances in the ambient
medium; such effects were simply not yet investigated. The
results discussed here, therefore, pertain only to the
accelerated particles.
Information on the proton and $\alpha$ particle abundances
relative to those of C and heavier nuclei could be obtained
from observations of the shapes of the gamma ray lines. In
Fig.~2 we show theoretical gamma ray spectra for the CRS
(cosmic ray source) and GR (grain) compositions calculated
using Eq.~2 (flat spectrum) with $E_c$=20 MeV/nucl. (Spectra
obtained with the strong shock spectrum, Eq.~(1), are not very
different.) The CRS spectrum (top panel) shows both broad and
narrow lines, as well as very narrow lines from long lived
radionuclei. The $^{12}$C complex around 4.44 MeV clearly shows
the narrow line superimposed on its broad counterpart. On the
other hand, the narrow lines are absent in the GR spectrum
(bottom panel) showing the effects of the absence of protons
and $\alpha$ particles. However, there are still very narrow
lines from the long lived radionuclei. To allow the shortest
lived radionucleus $^{55}$Co($\tau_{1/2}=17.5$h) to stop before
it decays, we assumed that the ambient density exceeds $2
\times 10^6$ cm$^{-3}$. By convolving spectra similar to those
in Fig.~2 with Gaussians representing the COMPTEL energy
resolution, Ramaty et al. (1995a) showed that the current data
cannot yet rule out a mixed broad-narrow line spectrum (e.g.
the CRS) even though the COMPTEL energy resolution at 4.44 MeV
(FWHM$\simeq$300 keV, Sch\"onfelder et al. 1993) would appear
sufficient. A similar result was also obtained by Cowsik \&
Friedlander (1995).
\begin{figure}
\picplace{11.3 cm}
\caption{Deposited power that accompanies the production of the
observed gamma ray line emission from Orion for the various
compositions and energy spectra discussed in the text. The
ambient medium is neutral and has solar system composition.}
\end{figure}
On the other hand, arguments of energetics tend to support an
accelerated particle composition poor in protons and $\alpha$
particles. To demonstrate this, we calculate the deposited power
that is associated with the production of the nuclear reactions
in a thick target. It is given by
\begin{eqnarray}
{dW \over dt} = \sum_{i}
A_i \int_0^{\infty} E {dN_i \over dt}(E) dE~,
\end{eqnarray}
\noindent where the $A_i$ is atomic number and ${dN_i \over
dt}$ is given by Eqs.~1 or 2. We emphasize that in a thick
target, the ratio of the nuclear reaction rate to the deposited
power is essentially independent of details of the interaction
region, especially the density of the ambient gas. It depends
mainly on the spectrum and composition of the accelerated
particles. The composition of the ambient gas could play a
major role but only if there were a substantial enhancement of
the C and heavier element abundances relative to H and He. Such
an enhancement would reduce the deposited power. There is also
a weak dependence on the state of ionization of the ambient
gas, as the Coulomb energy losses are higher in a plasma (by
about a factor of 2) than in a neutral gas. The results are
shown in Fig.~3 for an ambient neutral medium. The top panel is
for the strong shock spectrum (Eq.~1) and the bottom panel is
for the flat spectrum (Eq.~2). The deposited power, normalized
to the distance of Orion and the observed 3--7 MeV nuclear
deexcitation flux (Bloemen et al. 1994), is shown as a function
of $E_0$ or $E_c$ for the various compositions. We see that the
gamma ray line production is energetically most efficient for
compositions that are poor in protons and $\alpha$ particles
(i.e. the SN60, WC and GR compositions), favoring such
compositions. However, as we shall see below, even in the most
favorable cases, the rate of ionization of gas in Orion is
extremely large.
We mention here that if the light isotopes are indeed produced
mostly by low energy cosmic rays, then these cosmic rays must
also be depleted in protons and $\alpha$ particles. The
$\alpha$ particle depletion is necessary in order not to
overproduce $^6$Li; the proton depletion ensures a linear
dependence of the Be and B abundances on the Fe abundance in
stars of various ages (Duncan 1995). If the low energy cosmic
rays are poor in protons and $\alpha$ particles they will
produce Be and B from the breakup of accelerated C and O on
ambient H and He; in this case both the target and projectile
abundances could remain constant, leading to a linear growth of
the Be and B abundances. On the other hand, the GCR would
produce much of the isotopes from the breakup of C and O in the
ambient medium whose abundances increase with time, leading to
a quadratic growth.
Information on the composition of the heavy nuclei can be
obtained from the observed gamma ray spectrum. In Fig.~4 we
show calculations of the ratio $R$,
\begin{eqnarray}
R = 2 {\int_{1{\rm MeV}}^{3{\rm MeV}} E_{\gamma}^2
Q(E_{\gamma})
dE_{\gamma} \over \int_{3{\rm MeV}}^{7{\rm MeV}} E_{\gamma}^2
Q(E_{\gamma}) dE_{\gamma}},
\end{eqnarray}
\noindent for which Bloemen et al. (1994) set a 2$\sigma$ upper
limit of 0.13 (see also Fig. 1). We see that the CRS and SS
predicted ratios are in disagreement with the observations by
more than $3\sigma$, and the SN35 and SN60 predictions are
inconsistent at greater than $2\sigma$. Modifications in the
abundances, however, can invalidate these results. For example,
for the SN60 case, by increasing the C abundance by a factor of
2, but leaving all the other abundances unchanged, we obtain
$R\simeq 0.13$ for both $E_0$ or $E_c$ equal 30 MeV/nucl. On
the other hand, both the GR and WC compositions yield $R$'s
which are lower than the COMPTEL upper limit. However, while
the WC composition predicts practically no emission in the 1--3
MeV region, the GR composition predicts significant broad line
emission in this region, due to Mg, Si and Fe. The reduction in
the overall 1--3 MeV emission for the GR case is caused by the
absence of the 1.634 MeV line due to the lack of Ne in grains.
\begin{figure}
\picplace{11.4 cm} \caption{Calculated values for the ratio $R$
defined in Eq.~(4) for various compositions. The upper limit
established with COMPTEL is plotted at an approximate energy
which satisfies the constraints imposed by energetics, line
shapes and high energy gamma ray observations.}
\end{figure}
\subsection{Energy Spectrum}
We have already pointed out that the fact that the EGRET data
do not require an enhancement over the locally observed GCR
implies that the accelerated particles which produce the gamma
ray line emission in Orion are confined to low energies, below
the pion production threshold. On the other hand, the
energetics discussed above require a hard spectrum. This can be
seen from Fig.~3 which shows that for steep spectra (i.e. small
$E_0$ or $E_c$) the deposited power becomes very large. Thus
the combined COMPTEL--EGRET data imply that the low energy
cosmic rays in Orion should typically have energies around a
few tens of MeV/nucl.
The shapes of the observed gamma ray lines could place further
constraints on the hardness of the accelerated particle
spectra. For values of $E_0$ or $E_c$ larger than about 50
MeV/nucleon, for pure broad line spectra, the 4.44 and 6.13 MeV
lines become very broad (see Ramaty et al. 1996 for details).
Although the current COMPTEL data cannot yet rule out such
spectra, future observations could employ line shapes to limit
the hardness of the low energy cosmic rays.
\begin{figure}
\picplace{11.4 cm} \caption{Energy density in low energy cosmic
rays in Orion for various values of $E_0$ compared with the
energy density in the local Galactic cosmic rays.}
\end{figure}
\subsection{Low Energy Cosmic Ray Energy Density}
Whereas the power deposited in Orion by the accelerated
particles can be calculated independent of the total irradiated
mass of ambient gas, the low energy cosmic ray energy density
does depend on this total mass. We have calculated this energy
density assuming a steady state in which the accelerated
particles lose energy by Coulomb collisions in a neutral gas.
As already mentioned, the energy loss rates in an ionized gas
are higher by only a factor of 2.
The results are shown in Fig.~5, where $w(E)$ is the
differential energy density measured in MeV cm$^{-3}$
MeV$^{-1}$. The top (bottom) panel is for the CRS (WC)
composition. Also shown is the energy density in the local
Galactic cosmic rays (GCR) evaluated from direct cosmic ray
observations and assuming a power law in momentum source
spectrum (Skibo 1993). We see that for the CRS composition the
low energy cosmic rays energy density exceeds the local GCR
energy density by almost two orders of magnitude; for the WC
composition the excess is only about a factor of 10.
\subsection{Ionization of the Ambient Gas}
For the SN60, WC and GR compositions and $E_0$ or $E_c$ about
30 MeV/nucl, the deposited power is (2.5--5)x10$^{38}$ erg
s$^{-1}$. The total deposited energy depends on the duration of
the irradiation. For example, if the irradiation lasts 10$^{5}$
years, the total energy requirement would be
(0.8-1.6)x10$^{51}$ ergs, equal to the total kinetic energy
output of a supernova. Just such a supernova, occurring
$\sim$80,000 years ago in the OB association at $l$ =
208$^\circ$ and $b$ = $-$18$^\circ$, the same direction as the
center of the gamma ray line source, was suggested by Burrows
et al. (1993) from analy\-ses of the X-ray emission from the
Orion-Eridanus bubble (see also Ramaty, Kozlovsky \&
Lingenfelter 1996).
As it takes 36 eV of cosmic ray energy to produce an ion pair
in neutral H, the above deposited power corresponds to an
ionization rate of (4.3--8.7)$\times$10$^{48}$ H atoms s$^{-1}$
or (0.17--0.34) M$_\odot$ yr$^{-1}$. Even with no
recombination, in 10$^5$ years the amount of H that would be
ionized is only (1.7--3.4)$\times$10$^4$ M$_\odot$, which is
significantly smaller than the current neutral H mass in Orion.
Considering the current ionization rate per H atom, we obtain
that $\zeta$=(0.5--1.0)$\times$10$^{-13}$M$_5^{-1}$ s$^{-1}$,
where M$_5^{-1}$ is the irradiated neutral H mass in units of
10$^{5}$ M$_\odot$. The effects of such a very large ionization
rate have not yet been examined. It is nevertheless possible
that a large fraction of the power that accompanies the gamma
ray production is deposited in ionized gas.
\subsection{Origin of the Low Energy Cosmic Rays in Orion}
In considering the origin of the low energy cosmic rays in
Orion, we distinguish the problem of the acceleration from that
of the injection of the particles into the accelerator. The
proposed acceleration mechanisms are shock acceleration, due to
shocks associated with the winds of O and B stars (Nath \&
Biermann 1994) or shocks produced by colliding stellar winds
and supernova explosions (Bykov and Bloemen 1994), and
stochastic acceleration, due to gyroresonance with cascading
Alfv\'en turbulence in the accretion disk of a black hole
(Miller \& Dermer 1995). The proposed injection sources are the
winds of Wolf Rayet stars (the WC composition, Ramaty et al.
1995a), the ejecta of supernovae from massive star progenitors
(c.f. the SN60 composition, Cass\'e et al. 1995; Ramaty et al.
1996), and the pick up ions resulting from the breakup of
interstellar grains (c.f. the GR composition, Ramaty et al.
1995b;1996). The suppression of accelerated protons and
$\alpha$ particles relative to C and heavy nuclei (\S 3.1)
finds an explanation in stochastic acceleration. This
mechanism, however, does not predict the suppression of Ne and
heavier nuclei relative to C and O (\S 3.1). The suppression of
protons and $\alpha$ particles is relatively easily achieved by
all the proposed injection processes. Concerning the heavier
nuclei, Ne--Fe are strongly suppressed in WC, but less so in
the SN60. The suppression of Ne in the GR composition is
sufficient to account for the 1--3 MeV COMPTEL upper limit.
The comparison with solar flares is quite instructive (Ramaty
et al. 1995;1996). The solar flare gamma ray spectra show much
higher ratios of 1--3 MeV to 3--7 MeV fluxes than does Orion.
It was shown (Murphy et al. 1991) that this enhanced emission
below 3 MeV is, in part, due to the enrichment of the flare
accelerated particle population in heavy nuclei. Such
enrichments are routinely seen in direct observations of solar
energetic particles from impulsive flares (e.g. Reames, Meyer
\& von Rosenvinge 1994). These impulsive flare events are also
rich in relativistic electrons. On the other hand, in gradual
events the composition is coronal and the electron-to-proton
ratio is low. The acceleration in impulsive events is thought
to be due to gyroresonant interactions with plasma turbulence
while in gradual events it is the result of shock acceleration.
The fact that the ratio of bremsstrahlung-to-nuclear line
emission in Orion is very low lends support to the shock
acceleration scenario.
\section{Low Energy Cosmic Rays in the Galaxy}
We now provide estimates of the Galactic gamma ray line
emission expected from low energy cosmic ray interactions. We
base these estimates on the close relationship between light
isotope and nuclear deexcitation line production. In Fig.~6 we
show the B to 3--7 MeV nuclear gamma ray production ratio,
$Q({\rm B})/Q(3-7)$, as a function of $E_0$ and $E_c$ for the
strong shock (Eq.~1) and flat (Eq.~2) accelerated particle
spectra, and for the various compositions. We see that at the
higher values of $E_0$ and $E_c$ the ratio is not strongly
dependent on composition. This is because both the B and the
3--7 MeV deexcitation photons are mostly produced from C and O.
However, $Q({\rm B})/Q(3-7)$ does depend quite strongly on the
spectrum of the accelerated particles. For the subsequent
estimate, we take $Q({\rm B})/Q(3-7)\simeq$0.1, which, for
$E_0=E_c=20$ MeV/nucl, is a mean for the two assumed spectral
forms.
\begin{figure}
\picplace{11.7 cm} \caption{Production ratios of B to 3--7 MeV
nuclear deexcitation photons for various abundances and energy
spectra.}
\end{figure}
To estimate the total B inventory in the Galaxy we first use
the meteoritic B/H ratio given by Anders \& Grevesse (1989) and
a total Galactic mass of 5$\times$10$^{10}$ M$_\odot$. This
yields $N_{\rm B}$$\simeq$3$\times$10$^{58}$ atoms. However, as
the meteoritic B/H is probably higher than the B/H measured in
Pop I stars by a factor of 3 to 4 (Reeves 1994), we take the
Galactic B inventory at the formation of the solar system to be
approximately 10$^{58}$ atoms, and assume that it was produced
by low energy cosmic rays in about 5$\times$10$^9$ years. This
yields an average B production rate prior to the formation of
the solar system of about 6$\times$10$^{40}$ atoms s$^{-1}$. We
then further assume that current production rate is equal to
this average and use $Q({\rm B})/Q(3-7{\rm MeV})\simeq 0.1$ to
estimate a current Galactic 3--7 MeV photon production rate
$Q_{\rm G}(3-7{\rm MeV}) \simeq 6\times 10^{41}$ photons
s$^{-1}$. In terms of this total Galactic production, the flux
from the central Galactic radian is
\begin{eqnarray}
\Phi_{3-7}\simeq \xi 10^{-46} Q_{\rm G}(3-7) \simeq 6
\times 10^{-5}\xi~{\rm ph}~{\rm cm}^{-2}~ {\rm s}^{-1},
\end{eqnarray}
\noindent where $0.5 \lsim \xi \lsim 2$ depending on the
spatial distribution of the sources (Skibo 1993). Clearly this
prediction is highly uncertain. The estimate of the current B
production rate would be larger if there were significant
destruction of B due to incorporation into stars; in this case
we would predict a higher central radian flux. On the other
hand, the current B production rate could be lower than the
average rate prior to the formation of the solar system, in
which case our prediction would also be lower. In addition,
since the B to 3--7 MeV photon conversion depends on the
spectrum of the accelerated particles, for values of $E_0$ or
$E_c$ larger than 20 MeV/nucl, the ratio would be smaller than
the value we used, yielding a lower predicted central radian
flux. Clearly much could be learned from an actual measurement
of C and O deexcitation line emission from the direction of the
Galactic center or elsewhere.
In Fig.~7 we compare our calculations with observations. None
of these have revealed any line emission. The COMPTEL data
(Strong et al. 1994) is continuum, most likely a combination of
bremsstrahlung and inverse Compton radiation produced by
relativistic electrons. The SMM upper limit refers specifically
to line emission (Harris, Share, \& Messina 1995). We obtained
the calculated curves by normalizing the spectra shown in
Fig.~1 to the 3--7 MeV flux of $6 \times 10^{-5}~{\rm ph}~{\rm
cm}^{-2}~ {\rm s}^{-1}$ given above. While the calculations
are not inconsistent with data, it is conceivable that with
more sensitive instruments or with longer COMPTEL exposures,
the predicted line emission could be observed.
\begin{figure}
\picplace{5.9 cm} \caption{Predicted nuclear deexcitation line
emission from the central radian of the Galaxy. The COMPTEL
continuum observations (Strong et al. 1994) and the SMM upper
limit on line emission (Harris et al. 1995) are also shown.}
\end{figure}
For the WC, SN60 and GR compositions and $E_0 = E_c = 30$
MeV/nucl the energy required to produce 1 B atom is about 1 erg
(Ramaty et al. 1996). The current rate of B production
of about 6$\times$10$^{40}$ atoms s$^{-1}$ then implies a
current total power deposition by low energy cosmic rays in the
Galaxy of about 6$\times$10$^{40}$ erg s$^{-1}$. When compared
with the energy deposition rate in Orion (\S 3.4), we find
that about 200--400 Orion-like regions could be currently
active in the Galaxy. This small number, and the fact that the
irradiation time in Orion probably lasted for only 10$^{5}$
years, implies that the low energy Galactic cosmic ray
phenomenon is highly localized in space and time, in contrast
with the GCR whose spatial distribution is relatively uniform
and their time dependence, as evidenced by meteoritic studies,
is relatively constant in time.
\begin{acknowledgements}
I wish to thank S. W. Digel for discussions on the relationship
between the COMPTEL and EGRET data, and B. Kozlovsky and R. E.
Lingenfelter with whom much of this research was jointly
carried out.
\end{acknowledgements}
\begin{thebibliography}{}
\bibitem{} Adams, J. H. et al. 1991, ApJ, 375, L45
\bibitem{} Anders, E., \& Grevesse, N. 1989, Geochim. et Cosmochim.
Acta, 53, 197.
\bibitem{} Bloemen, H. et al. 1994, A\&A, 281, L5
\bibitem{} Burrows, D. N. et al. 1993, ApJ, 406, 97
\bibitem{} Bykov, A., \& Bloemen, H. 1994, A\&A, 283, L1
\bibitem{} Cass\'e, M., Lehoucq, R., \& Vangioni-Flam, E., 1995,
Nature, 373, 318
\bibitem{} Chaussidon, M., \& Robert, F. 1995, Nature, 374, 337
\bibitem{} Cowsik, R. \& Friedlander, M. 1995, ApJ, 444, L29
\bibitem{} Digel, S. W., Hunter, S. D., \& Mukherjee, R. 1995,
ApJ, 441, 270
\bibitem{} Duncan, D. 1995, paper presented at the Cosmic
Abundance Conference, College Park, MD, October 1995
\bibitem{} Harris, M. J., Share, G. J., \& Messina, D. C. 1995,
ApJ, 448, 157
\bibitem{} Ellison, D. C., \& Ramaty, R. 1985. ApJ, 298, 400
\bibitem{} Fisk, L. A., Kozlovsky, B., \& Ramaty, R. 1974, ApJ,
190, L35
\bibitem{} Ip, W.-H. 1995, A\&A, 300, 283
\bibitem{} Lingenfelter, R. E., \& Ramaty, R. 1976, ApJ, 211, L19
\bibitem{} Meneguzzi, M., Audouze, J., and Reeves, H. 1971, A\&A, 15, 337
\bibitem{} Miller, J. A., \& Dermer, C. D. 1995, A\&A, 298, L13
\bibitem{} Mitler, H. E. 1972, Astrophys. \& Sp. Sci., 17, 186
\bibitem{} Murphy, R. J., Ramaty, R., Kozlovsky, B., \& Reames, D.
V. 1991, ApJ, 371, 793
\bibitem{} Nath, B. B., and Biermann, P. 1994, MNRAS, 270, L33
\bibitem{} Ramaty, R. 1995, in The Gamma Ray Sky with COMPTON GRO and
SIGMA, eds. M. Signore, P. Salati, and G. Vedrenne, (Dordrecht:
Kluwer), 279
\bibitem{} Ramaty, R., Kozlovsky, B., \& Lingenfelter, R. E. 1979,
ApJ Suppl., 40, 487
\bibitem{} Ramaty, R., Kozlovsky, B., \& Lingenfelter, R. E. 1995a,
ApJ, 438, L 21
\bibitem{} Ramaty, R., Kozlovsky, B., \& Lingenfelter, R. E.
1995b, Ann. New York Academy of Sciences, (17th Texas Symposium
on Relativistic Astrophysics and Cosmology, eds. H. Bohringer,
G. E. Morfill and J. Trumper), 759, 392
\bibitem{} Ramaty, R., Kozlovsky, B., \& Lingenfelter, R. E.
1996, ApJ, 456, 525
\bibitem{} Reames, D. V., Meyer, J-P., \& von Rosenvinge, T. T.
1994, ApJ Suppl., 90, 649
\bibitem{} Reeves, H. 1994, Revs. Modern Physics, 66, 193
\bibitem{} Reeves, H., Fowler, W. A., \& Hoyle, F. 1970, Nature,
Phys. Sci, 226, 727
\bibitem{} Sch\"onfelder, V. et al. 1993, ApJ Suppl., 86, 657
\bibitem{} Skibo, J. G. 1993, Ph.D. Thesis, University of
Maryland
\bibitem{} Strong, A. W. 1994, A\&A, 292, 82
\bibitem{} Sofia, U. J., Cardelli, J. A., \& Savage. B. D. 1994, ApJ,
430, 650
\bibitem{} Woosley, S. E., Hartmann, D., Hoffman, R., D., \& Haxton,
W. C. 1990, ApJ ,356, 272
\end{thebibliography}
\end{document}