% http://www.oan.es/preprints
\documentstyle[epsfig, twocolumn]{iso98}
 
\setcounter{page}{1} 
%\newcommand{\thepapername}{}
 
%% Put any additional command definitions here
\newcommand{\RON}{read-out noise}
 
\begin{document}
 
%% Do not remove the following six lines:
\setlength{\parindent}{0pt}
\setlength{\parskip}{ 10pt plus 1pt minus 1pt}
\setlength{\hoffset}{-1.5truecm}
\setlength{\textwidth}{ 17.1truecm }
\setlength{\columnsep}{1truecm }
\setlength{\columnseprule}{0pt}
\setlength{\headheight}{12pt}
\setlength{\headsep}{20pt}
%% 
\pagestyle{veniceheadings}
 
%% Title - should be in capitals:
\title{\bf THE HOT GAS AND THE COLD DUST IN THE GALACTIC CENTER CLOUDS AS SEEN BY ISO \thanks{ISO is an ESA
project with instruments funded by ESA Member States (especially the PI
countries: France, Germany, the Netherlands and the United Kingdom) and
with the participation of ISAS and NASA.}}
 
 
%% If the author list spans more than one line then the {\bf (bold
%% font)} command must be inserted for each line
\author{{\bf J.~Mart\'{\i}n-Pintado$^1$, N.J. ~Rodr\'{\i}guez-Fern\'andez$^1$, 
P. ~de Vicente$^1$, A. ~Fuente$^1$,} 
\\ {\bf T.L. ~Wilson$^2$, S. ~H\"uttemeister$^3$, D. ~Kunze$^4$}\vspace{2mm} \\
$^1$Observatorio Astr\'onomico Nacional, Apartado 1143, E-28800, Alcal\'a de Henares, Spain \\
$^2$Max--Planck Institut fuer Radioastronomie, Postfach 2024, D 53010 Bonn,  Germany \\
$^3$Astronomische Institute, Auf dem Huegel 71, D-53121 Bonn 1, Germany \\
$^4$MPE, Giesenbachstrasse 1, D-85748 Garching, Germany. }

\maketitle
 
 
\begin{abstract}

We present the first results of the LWS and SWS ISO observations of 18 molecular
 clouds  in the Galactic center (GC) region.
%Towards all molecular clouds we find a complex structure with at least three
%different components: the ionized gas, the hot molecular gas and the cold dust
%and gas. The ionized gas show  we detect several  ^M
H$_2$ emission from several rotational lines in the ground vibrational state has been
detected  toward all molecular clouds. The rotational temperatures derived from
these lines
are between 150 and 500 K. Remarkably, the derived column densities of the hot
H$_2$, $\sim$ 2 10$^{22}$~cm$^{-2}$,  are similar to those obtained for the cold
gas.
From the non-detection of the S(3) line at 9.7\,$\mu$m, we find that the hot
gas must be located behind  more than 30 mag of visual extinction of cold gas
and dust. The  LWS continuum spectra  indicate that the cold  dust has
temperatures  of 25-35 K and column densities similar to those  required to
explain the extinction of the H$_2$ emission. Observations of the J=2-1 and
J=1-0 lines of C$^{18}$O  (J=2-1/J=1-0 line ratios of 0.7-1.4) seem to sample only the
cold  gas located in front of the hot gas.
This indicates that the low J
transitions of C$^{18}$O does not trace the total molecular gas in the GC. 
The LWS
continuum spectra impose stringent limits ($\leq$ 1 per~cent) to the hot 
dust associated to the hot H$_2$.


Toward most of the clouds we also detected emission from the fine structure
lines of
ionized  species such as  SIII and  NeII, and in some cases lines from NeIII and
OIII. In contrast with the previous idea that the heating of the GC clouds is
dominated by shocks, our data suggest that photoelectric heating 
by UV radiation can
be the dominat mechanism. The effective  temperature of the ionization radiation
derived from the NeIII/NeII
ratio is 35.000 K (typical of an O7 star). Our ISO data when combined
with the upper limits to the intensities of the recombination lines
measured at millimeter
wavelengths  indicate that the UV radiation is very diluted. Cavities with sizes
larger than 2 pc surrounding the ionizing  star(s) are required to explain the
 dilution. The origin of the ionized cavities
 and  the  implications  of these findings on the heating of the molecular
clouds in  the GC are  briefly discussed.    

 
 
\vspace{5pt} \\
 
%% Do not remove the previous commands. Your abstract should 
%% end with \vspace {5pt} \\  
 
 
%% Please insert your keywords here.
  Key~words: ISO; infrared astronomy; Galactic Center; ISM; shock waves.
 
 
\end{abstract}
 
\section{INTRODUCTION}
 
The molecular clouds in the GC exhibit  unusual characteristics  with  
widespread high gas kinetic temperatures and relatively low dust 
temperatures (see e.g. \cite{gus89}). Unlike the warm and hot cores  in the molecular 
clouds of the Galactic disk where the UV photons from young stars heat 
the grains and the gas is heated by  dust-gas collisions, the low dust  
temperatures observed in the GC require a different  heating mechanism 
which acts  directly on the gas. So far, the main heating mechanism invoked 
to explain the particular properties of the molecular clouds in the GC is  
shocks (\cite{wil82}). The chemistry in the GC clouds also seems  to be unusual.
 The abundance of refractory molecules like SiO is relatively large, 
supporting the shock scenario  (\cite{mart97}; \cite{hutt98}). 
Due to the large extinction towards 
this region  most of our view of the GC clouds has been obtained from radio 
observations. Obviously, a complete picture of the energetics of these     
clouds will only be obtained when the main cooling lines in the IR are 
measured.  We have used ISO to obtain a complete view of the ionized, hot 
and cold gas and dust in the GC clouds. These data will allow 
to fully characterize the GC clouds, and will help to understand 
the heating in the GC.
 
 
\section{OBSERVATIONS, DATA REDUCTION AND RESULTS}
 
We have observed 18 molecular clouds in the GC selected from the 
sample of \cite*{hutt93} and \cite*{mart97} with the LWS and the SWS of ISO. 
We have measured  with the SWS several rotational line of 
H$_2$ (S(0), S(1), S(3) , S(4) and S(5)), the 18.7 and 33.5\,$\mu$m 
lines of SIII, the  12.8\,$\mu$m of NeII,  and the 15.6\,$\mu$m line of NeIII. 
The reduction of the SWS data was  carried out with the Interactive 
Analysis programs at the ISO Spectrometer Data Centre at MPE. 
%Figure 1 and 2 
%show, respectively,  the ionized lines,  and H$_2$ lines towards M+0.16-0.1. 
We also took full LWS spectra of all sources. 
% and Figure 3 shows the spectrum towards M+0.16-0.1. 
Observations of the recombination line H41$\alpha$ and 
of the J=1-0 and J=2-1 lines of C$^{18}$O have also been obtained with an 
angular resolution similar to that of the SWS aperture using the IRAM 30-m 
telescope.
   \begin{figure}[h]
   \begin{center}
   \leavevmode
   \vspace{-1.7cm}
   \centerline{\epsfig{file=martinpj_1.eps, width=8.0cm}}
  \vspace{-1cm}
   \end{center}
 \caption{\em Fine structure lines observed with the SWS towards the
 cloud M+0.16-0.10.  The spectra are averaged to one half of the
 instrument's  resolution.}
 \label{fig:ionlines}
 \end{figure}    
 
\section{THE STRUCTURE IN THE GC CLOUDS}
%\label{sec:commands}
From the combination of  all these data, a new picture of the GC clouds emerges. 
We find that all the sources in our sample show the presence of three different components: the ionized gas, the hot molecular gas and the cold gas and dust. 

 
\subsection{The ionized gas}
%\label{sec:example}

Basically all the sources observed in the GC show emission 
from fine structure lines of ionized species like NeII and SIII (see Fig. 1). 
The ratio  between the 18.7 and the 33.5\,$\mu$m lines of SIII are 
$\leq$0.2. This is much smaller than the lower limit of 0.5 obtained from the 
excitation at very low density. This low ratio indicates that the 
ionized gas must 
be located behind more than 40 mag of visual extinction even in the case 
that the source of the ionized emission is extended. 
In the sources with the strongest 
NeII line,  we also detect emission from the NeIII line. For those sources 
the ratios between the NeIII and the 
NeII lines, 0.05-0.2, suggest effective temperature for the ionization 
radiation of at least 35000 K (\cite{sell96}),  typical of an O7 star. 
This is in contrast with the recombination line data  at millimeter 
wavelengths, for which we find an upper limit to the emission of 
$\sim$ 15 mK. Assuming the typical linewidths of $\sim$ 30  kms$^{-1}$, 
and a LTE electron temperature of 10000 K, our upper limits indicates 
that the Lymann continuum photons in the SWS aperture must be 
$\leq$4 10$^{47}$ s$^{-1}$, typical of a B0 star. The effective temperature 
of the ionization radiation derived from the NeIII/NeII ratio is 
inconsistent with ionization by main sequence OB starts exciting HII 
regions with sizes of $\leq$2 pc. As discussed in Section 4, the 
presence of large cavities surrounding the ionizing stars can explain 
the discrepancies. 
\begin{figure}[h]
   \begin{center}
   \leavevmode
   \vspace{-1.6cm}
   \centerline{\epsfig{file=martinpj_2.eps, width=8.0cm}}
  \vspace{-1.5cm}
   \end{center}
 \caption{\em Pure rotational transitions of H$_2$ as observed with the
SWS towards M+0.16-0.10. The spectra are averaged to one half of the
instrument's resolution. Note that the S(3) line is not detected
because of the absorption by the 9.7 $\mu$m silicates feature.}
 \label{fig:h2lines}
 \end{figure}      

 
\subsection{ The hot molecular gas}
%\label{sec:example}

 
 
All the clouds show emission in several rotational lines in the ground 
vibrational state of H$_2$ (see Fig. 2). In all sources the S(3) at 
9.7\,$\mu$m is weaker than expected, indicating  that  the hot gas is also 
behind of more than 30 mag of visual extinction. The rotational 
temperatures derived from the line intensity ratios 
vary from 150 K (S(1)/S(0)) to  500 K (S(5)/S(4)) (see Fig. 3). This indicates 
the  presence of large temperature gradients within the hot gas. 
Ortho-to-para ratios different from LTE will only change the 
kinetic temperature from 150 to 180~K.  

 \begin{figure}[h]
   \begin{center}
   \leavevmode
   \vspace{-1.2cm}
   \centerline{\epsfig{file=martinpj_3.eps,angle=-90, width=8.0cm}}
   \vspace{-1cm}
   \end{center}
\caption{\em Rotational plot for H$_2$ rotational states J=2-7 
obtained for M+0.16-0.10. 
%The logarithm
%of N$_j$/(g$_j$g$_s$) is plotted against the upper state energy in Kelvins,
%where N$_j$ is beam-averaged column density,
%g$_j$ is the rotational degeneracy, g$_s$ is the spin degeneracy.
The  filled squares have been obtained after correcting the 
measured values (open squares)
for 29 magnitudes of visual extinction using the extinction law 
of Draine 1989.
 The assumed ortho-to-para ratio is 3. Note that the values for
the S(3) line are upper limits.}
 \label{fig:grbolt}
 \end{figure}     

The total column densities of 
hot (150 K)  H$_2$ corrected by extinction  are at least 
1-2 10$^{22}$ cm$^{-2}$.  These column densities of hot molecular gas 
must be considered as lower limits since the lowest kinetic temperatures 
sampled by the S(0) and S(1) lines  are  $\geq$~140 K. If warm gas with kinetic 
temperatures of $\sim$~80 K is present, the column densities of the 
hot gas must be increased by a factor of more than  5. 
 
 
\subsection{ The cold molecular gas and dust}
%\label{sec:example}

The line ratios between the J=2-1 and the J=1-0 lines of C$^{18}$O 
towards all ISO sources are 0.7-1.4.  This seems to be a general 
characteristic in the GC molecular clouds (see e.g. \cite{hutt98}) and 
indicates that the molecular gas sampled by these lines  is relatively 
cold with kinetic temperatures much lower than the 140 K derived from the 
H$_2$ lines. The total H$_2$  column densities derived from the 
C$^{18}$O emission towards the ISO clouds are 3-5 10$^{22}$ cm$^{-2}$. 
These column densities are similar to those of the warm and the hot gas, 
but the measured line ratios of C$^{18}$O are inconsistent with those 
expected for the hot gas, namely  3-4. This suggests that the C$^{18}$O emission 
only samples the cold component. It is interesting to note that 
the column density of the hot component can be even larger 
than the column densities of the cold component measured from the 
C$^{18}$O lines. 

As mentioned in the previous section, the hot H$_2$ is located behind 
more than 30  mag of visual extinction. This material must be associated 
with relatively cold dust since it is outside of the region emitting 
the hot H$_2$. The  H$_2$ column densities associated with the 
material in front of the hot gas are similar to the estimated from 
the C$^{18}$O emission. This suggests  that both are associated. Further 
support for this idea comes from the spectrum measured with  LWS. The 
continuum spectrum  
between 40 and 190\,$\mu$m indicates that  most of the luminosity in the 
FIR is emitted by cold dust with temperatures between 25 and 35 K, and 
H$_2$ column densities of 3-8 10$^{22}$ cm$^{-2}$. The column densities 
and the temperatures derived from the dust emission are  consistent 
with those derived from the low J C$^{18}$O data. 

 \begin{figure}[h]
   \begin{center}
   \leavevmode
   \vspace{-1.2cm}
   \centerline{\epsfig{file=martinpj_4.eps,angle=-90, width=8.0cm}}
  \vspace{-1cm}
   \end{center}
 \caption{\em LWS spectrum of M+0.16-0.10. 
%There has been superimposed 
%two grey-bodies and its sum. 
%They show a cold (33 k) dust component of which corresponds
%to an H$_2$ column density of 2.7 10$^{22}$ cm$^{-2}$ (dashed line), 
%a warm (120 K) component (solid line) with H$_2$ column 
%density 1000 times smaller
%than the  cold component  that is being attenuated  by the cold dust, 
%and the sum of the cold and the warm grey-bodies (squares).}
%
%There has been superimposed a cold (33 k) dust component of N$_{H_2}$ $\sim$
%3 10$^{22}$ cm$^{-2}$, a warm (120 K) component (solid line) with
%N$_{H_2}$ 1000 times smaller that is being attenuated  by the cold dust,
%and the sum of the cold and the warm grey-bodies (squares).} 
%
It    has been superimposed a two component grey-body (squares) where a 
cold (33 K) dust component (dashed line)  
of N$_{H_2}$ $\sim$ 3 10$^{22}$ cm$^{-2}$ 
is attenuating a   warm (120 K) dust component with N$_{H_2}$ 1000 
times smaller (solid line). Note that even such an small amount of hot dust
should have been detected.} 


 \label{fig:lws}
 \end{figure}                    


From the continuum data at short wavelengths, we derive an upper limit 
to the column density of the hot dust with similar temperatures to that of  
the hot gas  of $\leq$ 10$^{20}$ cm$^{-2}$, more than 100 times
smaller than that obtained from the H$_2$ lines (see Fig. 4). Our data 
support the  idea that the dust associated with the hot gas 
is much colder than the gas. 


\section{DISCUSSION }
\label{sec:commands}

Since the discovery of the relatively uniform high temperatures in the
envelope  of Sgr B2 and in other molecular clouds in the GC, 
a detailed understanding of the thermal balance in these molecular 
clouds has been a challenge. The most accepted idea is that shock heating 
explains  the large column densities of hot gas associated with relatively cold dust (\cite{wil82};\cite{gus89}). 

\subsection{The heating mechanisms}
%\label{sec:example}

Our ISO data show that the material along the line of sight 
towards the GC molecular 
clouds has  a complex structure.  The presence of an ionized component 
requiring an effective temperature for the ionization radiation of 
$\sim$35000 K indicates that shock heating is not the only mechanism 
which heats the large column densities of hot molecular gas in the GC clouds. 
If  shocks are responsible for all of the heating, the presence of the 
NeII and NeIII lines would imply J type shocks with shock velocities 
$\geq$ 100 kms$^{-1}$. This is contradiction with the profiles of the 
ammonia lines arising from the hot and the cold gas (see \cite{hutt93}) 
which do not show the displacement in radial velocities expected if 
the hot ammonia were heated by J-shocks. Furthermore, the ratio of 
the total      
line cooling rate  to the FIR dust continuum is  typically  0.006, 
far too low for the expected values from fast J shocks (\cite{holl89}).
 
On the other hand,  photoelectric  heating will account for most of the 
observed characteristics of the ionized and the hot molecular components. 
For typical photodissociation regions (PDRs) in the Galaxy, photoelectric 
heating heats the gas to temperatures above those of the dust in a region 
with typical column densities of $\sim$ 3 10$^{21}$ cm$^{-2}$ 
(see \cite{holl97}). Several typical Galactic PDRs along the line of 
sight are then required to account for the large H$_2$ column densities 
of hot gas observed with ISO. 
 
\subsection {Wind blown ionized bubbles } 
%\label{sec:example}

Further support for the presence of several PDRs along the line sight 
of the GC comes from the inconsistencies between   the  Lyman continuum 
photons and  the effective temperature of the ionization radiation. 
This can be reconciled if the ionized gas fills relatively large 
cavities surrounding the ionizing stars.  Assuming a  main sequence O star, 
our upper limit to the recombination line emission sets a lower limit 
to the size of the cavities of 2.5 pc.
Hot (60-130 K) shells of molecular gas with sizes of 1-2.5 pc have 
recently been observed in the envelope of  Sgr B2 molecular 
cloud (\cite{mart99}), and approximately 300 shells with sizes 
between 5 and 10 pc have been also identified from the CO survey 
of the GC (\cite{hase98}). From the number and the sizes of the 
shells detected in the GC, it is very likely that basically any line 
of sight towards the GC intersect at least two shells. In fact,
 \cite*{hase98} claim that the most likely occurrence of shells along 
the line of sight is $\sim$ 20 shells. This would easily explain that 
the large column densities of hot H$_2$ observed with ISO can be 
produced by PDRs at the inner edges of a number of shells.
 
Although photoelectric heating can explain the ISO observations, 
the presence of a relatively large column density of hot ammonia 
($\sim$ 30 per~cent of the cold ammonia, see \cite{hutt93}) indicates 
that the Far UV radiation in not the only  heating mechanism acting in 
the GC clouds. This is due to the fact that ammonia is a very 
fragile molecule and would be easily   photo-dissociated  in typical 
PDRs (\cite{fuen90}). \cite*{mart99} have found that the hot ammonia 
shells in the Sgr B2 molecular cloud are expanding at relatively low 
velocities, 6-10 kms$^{-1}$. They proposed that the hot ammonia 
shells are produced  by  shocks associated with  wind blown bubbles 
driven by evolved massive stars such us  Wolf-Rayet stars or Luminous blue 
variables. These wind blown bubbles can 
explain  the large cavities  around 
the stars, the rather high effective temperature of the ionization radiation,  and the shock chemistry in the GC molecular clouds.

In summary, the ISO data show that the energetic of the GC molecular clouds
is dominated by FUV radiation that ionizes and heats the gas 
by photoelectric effect  and to less extent by 
low velocity shocks, caused by the wind of evolved massive stars, that 
contribute to the selective heating of the gas maintaining the dust at lower
temperatures.

 
\section*{ACKNOWLEDGMENTS}
We acknowledge support from the ISO Spectrometer Data Centre at MPE Garching, 
funded by DARA under grant 50 QI 9402 3. JM-P,NJR-F, PdV and A.F have been 
partially supported by the CYCIT and the PNIE under grants PB96-104 and 
ESP97-1490-E. NJR-F acknowledges a grant from Conserjer\'{\i}a de 
Educaci\'on y Cultura de la Comunidad de Madrid. 
 
 
 
\begin{thebibliography}{}


\bibitem[\protect
\astroncite{Draine}{1989}]{draine89}
Draine B. T., 1989, in Proc. 22nd ESLAB Symp. on IR Spectroscopy in
Astronomy, ed. B.H.Kaldeich, ESA SP-290,93
 
\bibitem[\protect
\astroncite{Fuente  et~al.}{1990}]{fuen90}
Fuente, A., Mart\'{\i}n-Pintado, J., Bachiller, R., \& Cernicharo, 
J. 1990, A\&A, 237, 471
 
\bibitem[\protect
\astroncite{ G\"usten}{1989}]{gus89}
G\"usten, R. 1989, in IAU Symposium 136, The Galactic Center 
of the Galaxy, ed. M.  Morris, Kluwer Academic Publisher, p. 89 
 
\bibitem[\protect
\astroncite{ Hasegawa et~al. }{1998}]{hase98}
Hasegawa, T., Oka, T., Sato, F., Tsuboi, M., \& Miyasaky, A. 1998, in 3rd 
Cologne-Zermatt Symposium, The Physics and Chemistry of the Interstellar Medium, 
ed. V. Ossenkopf,  Shaker Verlag, in press 
 
\bibitem[\protect
\astroncite{ Hollenbach \& McKee }{1989}]{holl89}
Hollenbach, D.,  \& McKee, C.F. 1989, ApJ. 342, 306
 
\bibitem[\protect
\astroncite{ Hollenbach \& Tielens }{1997}]{holl97}
Hollenbach, D.,  \& Tielens, A.G.M. 1997, ARA\&A. 35, 179
 
\bibitem[\protect
\astroncite{H\"uttemeister et~al.}{1998}]{hutt98}
H\"uttemeister, S., Dahmen, G., Mauersberger, R., Henkel,  C., 
Wilson, T. L., \&  Mart\'{\i}n-Pintado, J. 1998, A\&A, 334, 
 
\bibitem[\protect
\astroncite{H\"uttemeister et~al.}{1993}]{hutt93}
H\"uttemeister, S., Wilson, T. L., Bania, T. M., \&  
Mart\'{\i}n-Pintado, J. 1993, A\&A, 280, 255
 
\bibitem[\protect
\astroncite{ Mart\'{\i}n-Pintado et~al.}{1997}]{mart97}
Mart\'{\i}n-Pintado, J., de Vicente, P., Fuente, A., \& Planesas, P. 1997, 
ApJ, 482, L45
 
\bibitem[\protect
\astroncite{ Mart\'{\i}n-Pintado et~al.}{1999}]{mart99}
Mart\'{\i}n-Pintado, J., Gaume, R., Rodr\'{\i}guez-Fern\'andez, N.J., 
de Vicente, P., \& Wilson T.L. 1999,  ApJ, in press

\bibitem[\protect
\astroncite{Sellmaier et~al.}{1996}]{sell96}
Sellmaier, F.H., Yamamoto, T., Pauldrach, A.W.A., \& Rubin, R.H. 1996, A\&A, 305, L37 

\bibitem[\protect
\astroncite{ Wilson et~al.}{1982}]{wil82}
Wilson, T.L., Ruf, K., Walmsley,  C.M., Martin, R.N., Pauls, T.A., \&  Batrla, W. 1982, A\&A, 115, 185
 
\end{thebibliography}
 
 
\end{document}