The Infrared Band Strengths of H2O, CO and CO2 in Laboratory
astro-ph/9409076 14 Mar 1995 in band is found to the sitions strengths of the CO CO ice strengths frared is imp ortan accuracies ice material Abstract man The analogs comp osition lab oratory P A L Dep band eiden of tles sho artment Gerakines the b etter infrared Impro t to w strengths Infrared W Observatory In molecule that e ha order of b e presen v than v of ed e go o d measuremen the H w Physics exp erimen eak sp ectroscopic to sim W of O a ice to t Accepted er than previously measured for few accurately CO and H the band CO b e P O matrix A Sc ulations p ercen results and studied R and ensselaer tal Box h O infrared features are found to dep end only w utte H b strengths and metho ds lik y observ t calculate O of Astr and ely ts of F the are lab oratory urthermore J M Green Polyte onomy ations of CO of rev studied their infrared band strengths in astroph the relying the ersible astroph chnic are RA to dominating column w exp erimen of imp ortan Astr ard L the In b erg on Institute temp erature eiden H con ophysics sim ob jects temp erature a mixture of densities ysical trast O ultaneous and E F v The t ts ice constituen T to obscured r CO with comp onen Netherlands oy dep endence determine of ice NY b eha previous these indep enden CO in H an and ts b Disho ec vior mixtures t y of molecules dense ha these in CO of w of O v USA terstellar eakly on ork e the t the k led dep o ysical cloud band CO the in to it
1. Introduction
Infrared astronomy has led to the identi cation of various sp ecies existing within
icy grain mantles in dense clouds The two most abundant molecules in grain mantles
identi ed to date are water H O and carb on monoxide CO e g Willner et al
Smith et al Whittet et al Whittet et al Tielens et al Chiar et
al Limited observational evidence as well as theoretical and lab oratory mo deling
indicate that carb on dioxide CO should also b e an imp ortant comp onent of interstellar
ices d Hendecourt Jourdain de Muizon Whittet et al Breukers The
abundances of these molecules in di erent phases of ice mantles provide imp ortant clues
to the chemical pro cesses in dense interstellar clouds and therefore it is of imp ortance to
accurately measure the band strengths of the infrared features of these molecules The goal
of this pap er is to provide new as well as more accurate measurements of the band strengths
of H O CO and CO contained within astrophysical ice analogs so that their abundances
in the ices in dense interstellar clouds may b e determined with more con dence Such
measurements are particularly imp ortant for CO a molecule whose principle absorption
band is totally obscured by Earth s atmosphere and which will b e widely searched for with
the Infrared Space Observatory ISO
Astrophysical ices consist of complex mixtures of molecules e g Whittet and
references therein At least two distinct phases app ear to exist one which is dominated
by p olar molecules of which the most abundant is H O and one dominated by ap olar
molecules such as CO e g Sandford et al Tielens et al It is not yet clear
which sp ecies dominate the ap olar phase although CO and p ossibly O and CO seem to
b e abundant Furthermore observations towards emb edded sources indicate a wide range
of ice temp eratures from less than K up to K Smith et al
Observations of the m cm C O stretching band indicate that CO re
sides in b oth the p olar and ap olar phases of interstellar ice Eiroa Ho dapp Kerr
et al Chiar et al CO could b e present in b oth phases as well Although
H O app ears to b e the most abundant molecule in the p olar phase the presence of ap olar
sp ecies like CO and CO could in uence the strength of its infrared bands Sp eci cally
the resulting break up of the hydrogen b onding network would weaken the intensity of
the O H stretching feature of fully H b onded H O at m cm Hagen Tielens
As of yet H O has not b een observed in the ap olar ice
Previous metho ds of measuring band strengths of molecules in a mixed ice involved
the preparation of several gases within one gas container by adding the comp onents one
after another and using the ideal gas law to convert pressures to abundances This mixture
is then dep osited onto a cold substrate and it is assumed that the molecular abundances
in the ice sample equal the gas abundances in the container The ice abundances are then
used to convert the measured integrated optical depths to infrared band strengths either
by measuring the thickness of the ice from the interference fringes pro duced by a laser
directed at the sample and assuming a value for the ice density thus obtaining molecular
column densities d Hendecourt Allamandola Hudgins et al or by assuming
that the band strengths of the dominant comp onent are equal to those in a pure sample and
using these to calibrate the other features Sandford et al Sandford Allamandola
These metho ds rely heavily on the assumed equality of the comp ositions of the
ice sample and the gas mixture However several problems may arise in this assumption
First if a bulb is made with H O close to its vap or pressure a small change in temp erature
could signi cantly in uence the amount of H O in the gas phase inside the bulb due to
the strong temp erature dep endence of the vap or pressure For example b etween and
C it varies from to mbar Second mixing of the di erent gases which are
sequentially allowed to enter the bulb may b e incomplete Finally the dep osition rate of a
molecule will b e prop ortional to its thermal velo city which dep ends on the molecular mass
1
2
This e ect could give rise to a signi cant di erence in the comp osition of i e v m
th
the gas mixture and the ice sample if molecules with very di erent molecular masses are
involved For example H O and CO have thermal velo cities that are di erent by a factor
of
In the new pro cedure that we have implemented we have avoided these pitfalls by
pro ducing ice mixtures using simultaneous dep ositions of pure gases through separate
dep osition tub es We then have the ability to measure the ratio of the band strengths of
molecules in a binary ice to those in pure ice A A since the column densities in each
pure
case pure and mixed can b e kept the same The band strengths in the mixed ice can
then b e calculated using the strengths of the pure ice bands which are accurately known
and present in the literature
This pap er is organized as follows In x we review our exp erimental metho ds In x
we summarize our measurements of the band strengths of H O CO and CO in various
binary mixtures with p olar and ap olar molecules Finally in x we discuss our results
and their astrophysical implications
2. Experimental
In this section we will describ e the exp erimental equipment and pro cedures used for
pro ducing and analyzing astrophysical ice analogs
2.1 Sample Chamber
The vacuum system used to prepare the ice samples and to obtain infrared transmis
sion sp ectra is similar to those previously used to study ice analogs Hagen et al
Allamandola et al Hudgins et al with some signi cant mo di cations The
chamb er is situated within the sample compartment of an infrared sp ectrometer Bio Rad
FTS A An infrared transmitting substrate CsI is mounted in the vacuum chamb er
and can b e co oled by a closed cycle helium refrigerator expander Air Pro ducts Displex
DE compressor mo dule Air Pro ducts R W SL to a temp erature of K The
temp erature of the substrate is continuously adjustable by a resistive typ e heater element
up to ro om temp erature The temp erature is monitored by a chromel Au thermo couple
with an accuracy of K The chamb er has four p orts Two of these consist of KBr al
lowing transmission of the infrared b eam of the sp ectrometer One of the p orts consists of
MgF to enable ultraviolet UV irradiation of the ice samples this option was not used
for the exp eriments describ ed in this pap er while the fourth p ort consists of glass and is
used for visual monitoring Additionally the chamb er is equipp ed with two dep osition
Figure Diagram of the substrate holder showing the cm o axis mounting of the
substrate a so that it may b e turned completely out of the infrared b eam b the b eam
is represented by the dotted line and the cross denotes the rotation axis
tub es which are directed at the center of the substrate from a distance of cm For
each dep osition system the gas ow from a storage bulb see x to the vacuum sys
tem is regulated by a variable leak valve Leyb old Heraeus The leak valve is
equipp ed with a shut o valve and a regulation valve that function indep endently The
stainless steel chamb er is sealed with Viton O rings and is evacuated by a turb o molecular
pump Pfei er Balzers TPH backed by a rotary pump Pfei er Balzers DUO B
A liquid nitrogen trap b etween the two pumps prevents any rotary pump oil from back
streaming to the system Ion and thermo couple pressure gauges placed b etween the sample
chamb er and the turb o pump monitor the internal pressure The vacuum system can b e
externally heated using heating tap e After ab out two days of pumping and externally
heating to C a vacuum of mbar was obtained The residual gases at this
pressure were analyzed by collecting them on the substrate after co oling to K and sub
sequent infrared sp ectroscopy It was found to consist mainly of H O accreting at a rate
of molec cm s m hr while organic contaminants were found to
accrete at ab out half this rate
The substrate holder can b e rotated without breaking the vacuum During an exp eri
ment the substrate is rotated b etween two p ositions which are degrees apart In p os
an infrared transmission sp ectrum of the substrate and the ice sample can b e obtained
The sp ot size of the infrared b eam on the CsI substrate is cm in diameter Position
enables dep osition of an ice sample through one or b oth dep osition tub es onto the cold sub
strate Also the ice sample can b e UV irradiated in this p osition Moreover the substrate
holder is constructed such that the substrate is mounted cm away from the rotation
axis Therefore while the substrate is in p os the infrared b eam may pass unimp eded
through the sample chamb er see Fig This allows the collection of a reference infrared
sp ectrum at any moment during the exp eriment Use of such reference sp ectra greatly
improves the baseline stability of the infrared sp ectra see x
The homogeneity of the ice samples pro duced with the two dep osition tub es was
checked by varying the sp ot size of the infrared b eam on the sample using the variable
ap erture in front of the infrared source of the sp ectrometer Samples were found to b e
homogeneous within inside a diameter of cm
2.2 Gas Bulb Preparation
Bulbs for individual gases were prepared in a glass vacuum manifold evacuated
through an oil di usion pump Edwards Di stack Series Mo del backed by a rotary
mechanical pump Edwards BS Back streaming of pump oil into the system is pre
vented by a liquid nitrogen cold trap lo cated b etween the line and the pumps The pressure
in the line can b e monitored in many ranges with an ion gauge a thermo couple gauge
and a diaphragm manometer The system is heated with heater tap e under vacuum to
clean it b etween preparation of di erent samples The comp ounds used and their purities
are as follows H O liquid triply distilled CO gas Messer Griesheim purity
CO gas medical grade purity O gas Messer Griesheim purity
Pressures used in bulbs of H O were limited to a maximum of mbar well b elow
its ro om temp erature vap or pressure of mbar We have thus avoided the e ects of
substantial sticking of H O on the walls of the bulb and the temp erature dep endence of
the H O vap or pressure as describ ed in x
2.3 Experimental Procedures
The exp erimental pro cedures applied for measuring the infrared band strengths for
a molecule in a binary ice are describ ed here Two gas bulbs one containing the gas for
which the infrared band strengths are to b e measured henceforth the sub ject gas and
one containing the gas with which this sp ecies is to b e diluted henceforth the dilutant
are connected to the entries of the two dep osition tub es Before co oling the substrate ow
rates are set with the regulation valve while monitoring the pressure increase within the
system using the following relation
P
F
p
i
m
i
where m is the molecular mass of comp onent i P is the measured pressure increase
i
within the vacuum system and F is the desired ow rate of comp onent i Schutte et
i
al The ow rates through each dep osition tub e may b e set indep endently in this
way based on the desired ratio of the comp onent gases After setting the ow rates with
the regulation valve ows are controlled exclusively with the shut o valve From later
sp ectroscopic analysis as describ ed in x of ices pro duced by this metho d of pre
calibration it is found that the ow rates repro duce within
After calibration the substrate is co oled down to K and a dep osition is made of
the sub ject gas After obtaining the infrared sp ectrum the substrate is heated until the
sample sublimes and then reco oled Next the sub ject gas and the dilutant are dep osited
simultaneously for the same length of time as the rst dep osition resulting in a binary ice
sample containing the same numb er of sub ject molecules The column density N of an ice
comp onent and the integrated optical depth of its absorption band i relate as
R
d
i
N
A
i
where A is the band strength and the integration is taken over the feature in question
i
using an appropriate baseline Thus the ratio of the strengths in the binary and in the
pure ice can b e found simply from the ratio of the integrated optical depths in the two
samples If the ratio of sub ject dilutant is large ab out an additional sample of
pure dilutant is made as a nal step Then the exact comp osition of the binary ice can
b e determined from the integrated optical depths measured in the two pure samples using
band strengths from the literature If the sub ject material makes up less than of
the binary ice the comp osition of the binary ice is simply determined from the integrated
optical depths of the dilutant in the binary ice and of the sub ject material in the pure ice
and no additional dep ositions are p erformed
Ice samples used in our exp eriments had the approximate comp osition dilutant sub ject
or Due to uncertainties involved in pre calibrating the ows using the in
duced pressure increase the actual comp osition of the ice sample as it was measured
from the infrared sp ectra could di er by up to a factor of two from what was intended
although the deviation was typically no more than The total dep osition rate
when pro ducing the binary ice was typically molec cm s for the
ices and molec cm s for the ices corresp onding to thickness growth
rates of and m hr resp ectively Dep osition rates of the sub ject gases were
molec cm s when pro ducing a ice sample and molec cm s
when pro ducing a ice sample With these dep osition rates the contamination due to
mixing of the residual gases into the ice samples ranged from to Dep osition times
were equal to min for the ices and min for the ices and hence the resulting
thicknesses of the ice samples were and m resp ectively Dep ositions at K
instead of K were made in order to check the sticking of the molecules on the substrate
i e if no decrease in sticking at the higher substrate temp erature is found then sticking
can b e assumed to equal Sandford Allamandola In all cases sticking was
found to equal unity
2.4 Measurement of Spectra
The sample chamb er is situated within the sample compartment of the infrared sp ec
trometer such that the infrared b eam axis is aligned with the center of the cold substrate
in p os Single b eam sp ectra were taken from to cm to m at a
resolution of cm the width of an unresolved line Pro ducing a sample sp ectrum
involved obtaining a single b eam sp ectrum of scans b efore and after dep osition with
subsequent ratioing see e g Hudgins et al
In some cases esp ecially when measuring sp ectra after gradual warm up of the ice
sample the time b etween obtaining the background and the sample may b ecome large
p erhaps on the order of hours Due to inevitable sp ectrometer instabilities such as in
the source temp erature and in the alignment of the interferometer this can lead to the
app earance of some spurious sp ectral structure and deviation of the sp ectral baseline from
zero absorbance To prevent this additional single b eam sp ectra can b e obtained just
after the background sp ectrum at time t and just after the sample sp ectrum at time
t with the substrate in p os i e with the infrared b eam passing unimp eded through
the vacuum chamb er an empty sp ectrum The sample absorbance sp ectrum is then
obtained from
Background t Sample t
log Abs log
Empty t Empty t
3. Results
3.1 Pure ices
Integrated optical depths in cm were measured from the absorbance sp ectra by
cho osing an appropriate baseline for the band in question For bands of pure ice and for
most ices without H O a linear baseline was used The bands of H O are strong and
broad and they dominate in most of the mixed ice sp ectra Most CO and CO bands
lie on top of these broad structures In order to de ne a go o d baseline in these cases
we have made p olynomial ts to the underlying structure and used those which closely
approximate its curvature In general go o d ts were obtained for p olynomials of order or
In some cases an average of measurements was taken of a band for which more than one
p olynomial t well An example of a band for which the underlying H O feature created
some di culty in the determination of the baseline is the CO cm m band
which lies on top of the H O m cm feature This band and two p olynomial
ts of the underlying H O feature are shown in Figure for an H O CO mixture
The di erence in measured integrated optical depth b etween the two cases is
Table lists the infrared band strengths for the pure ices used in these exp eriments
For each molecule values for the strongest features have b een taken from the literature
Yamada Person Jiang et al Hagen et al and used to calculate the
strengths for the other bands by scaling the relative integrated optical depths For the
bands of CO and CO a terrestrial isotopic ratio of C C has b een used to
further scale the band strengths Measurements of pure CO at higher temp eratures were
obtained by dep ositing the ice at K followed by stepwise heating of the ice sample at
a rate of ab out K min and taking an infrared sp ectrum at each step This pro cedure
reveals only a weak temp erature dep endence of the CO infrared bands Band p ositions
in b oth m and cm are listed for the pure ices cf Sandford et al Sandford
Allamandola For purp oses of identi cation we will denote each band by its
p osition in the pure ice Peak band p ositions in mixed ices were found to agree with those
of Sandford et al and Sandford Allamandola and we refer the reader to
these works for a complete listing of band p ositions in binary mixtures
Figure The cm m band of CO in an H O CO mixture demon
strates the uncertainty involved in pro ducing a baseline t due to the underlying feature
of H O Dotted line st order p olynomial t to the H O band dashed line nd order t
in the same region
3.2 CO mixtures
The band strengths for the C O and C O fundamental stretching mo des are
shown in Table for CO mixed with H O O and CO Errorbars were estimated from
the results obtained with various p olynomials used as a baseline t Only errors larger
than are listed Errors due to sp ectral noise are negligible i e less than even for
the weakest features In each case the value of the CO band intensity ratio ed by that of
pure CO do es not deviate from unity by more than The values for the CO band in
H O contain a substantial baseline error due to its small size and its p osition on a broad
H O feature Within the errorbars these values remain close to the pure ice value as well
This also holds for non H O ices since the value of the CO band strength app ears to
stay close to the value for the pure ice within the given uncertainty It must b e noted
that since O is infrared inactive the comp osition of the O containing binary ices could
only b e assessed from the increase in pressure at ro om temp erature by the O ow x
Thus for these ices there can b e an error in the listed comp osition of ab out a factor of
Table Infrared band intensities of pure H O CO and CO ices after dep osition at K
and CO after warm up to and K
A A A A A A A
K K K K K K
Ice Mode Band Position K
cm ( m) cm molec cm molec cm molec a
H O ...... O±H stretch 3280 (3.045) 2.0(-16) O±H bend 1660 (6.024) 1.2(-17)
libration 760 (13.16) 3.1(-17)
b
CO...... C O stretch 2139 (4.675) 1.1(-17)
C O stretch 2092 (4.780) 1.3(-17)
c
CO ...... ( ) C=O stretch 2343 (4.268) 7.6(-17) 0.98 7.4(-17) 0.97 7.4(-17)
( ) C=O stretch 2283 (4.380) 7.8(-17) 0.94 7.3(-17) 0.92 7.2(-17)
( ) O=C=O bend 660,665 (15.15,15.27) 1.1(-17) 1.03 1.1(-17) 1.04 1.1(-17)
( ) combination 3708 (2.697) 1.4(-18) 1.05 1.5(-18) 1.08 1.5(-18)
( ) combination 3600 (2.778) 4.5(-19) 1.22 5.5(-19) 1.20 5.4(-19)
b a b c
a(-b)=a ; Hagen et al. 1981; Jiang et al. 1975; Yamada & Person 1964
of course this do es not intro duce any uncertainty in the measurement of A A see
pure
x
We have studied the temp erature dep endence of the C O band strength in the
H O CO ice sample Measuring the irreversible comp onent of the temp erature
dep endence is imp ossible for CO since the weakly b ound fraction of the CO molecules
already starts to sublime when the sample temp erature is raised to K Schmitt et
al Sandford et al For this reason we have only measured the reversible
comp onent of the temp erature dep endence following the metho d of Schmitt et al
The sample was initially dep osited at K and then heated to K and allowed
to anneal for minutes to sublime any weakly b ound CO from the sample The sample
was then held at a temp erature of K for minutes in order to stop sublimation and
to pump away any gas phase CO left in the system The sample is then reco oled and
measured at K After annealing the ice comp osition was found to equal H O CO
when assuming that A CO K cm molec as for the unannealed
ice As a check of CO re condensation during this co oldown the sample was kept at K
for min and then re measured No new condensation was found in our ice sample The
sample is then warmed up in steps and a sp ectrum taken at each step In order to check
whether any further sublimation of CO o ccurred during this second warm up sequence the
sample was once again co oled to K and the CO band measured The intensity of the CO
band was found to remain constant relative to the initial annealed ice Figure shows the
obtained reversible temp erature dep endence of the C O fundamental stretching mo de
As in Schmitt Et al a clear temp erature dep endence is found Our results show
that the band strength drops by when the temp erature is raised from to K
Table Infrared band intensities of CO in binary ices with H O O and CO after
dep osition at K Uncertainties greater than are listed
Ice Mo de A A A
pure
cm molec
H O CO C O stretch
a
C O stretch
H O CO C O stretch
a
C O stretch
b
O CO C O stretch
C O stretch
b
O CO C O stretch
C O stretch
CO CO C O stretch
C O stretch
CO CO C O stretch
C O stretch
b a
a b a Measurement of integrated optical depth contains
b
a large uncertainty due to underlying features of H O Listed
comp osition is uncertain by a factor of see text
3.3 CO mixtures
Table presents the band strengths for the infrared bands of CO and the strongest
band of CO in H O O and CO mixtures Errors due to baseline uncertainties are listed
whenever they exceed as in Table We have investigated the temp erature dep endence
of the CO bands in an H O matrix by warm up of the sample following dep osition as
discussed ab ove for pure CO see x No signi cant temp erature dep endence was
found at K the strengths of all CO bands measured deviated by less than from
their values at K As with CO the CO band strengths dep end only slightly on the
ice comp osition The strength of the main stretching feature of CO varies by less than
for the ices used here The strengths of the CO band and the b ending mo de of
CO seem to deviate by a somewhat larger amount up to and resp ectively but
these measurements can contain considerable error up to Only the strengths of
the two weak CO combination bands de nitely show considerable dep endence on the ice
comp osition
In order to estimate the in uence of residual gases dep ositing within the ice samples
an H O CO exp eriment was p erformed with a times higher dep osition velo city
Figure The reversible temp erature dep endence of the CO infrared band strength in an
H O dominated matrix H O CO The absolute scale shown on the right y axis is
obtained by assuming A CO K cm molec as in the ice sample b efore
annealing
The obtained pure to binary ratios deviated by less than from those obtained at the
lower dep osition rate indicating that the residual gas condensation do es not a ect the results
3.4 H O mixtures
Figures and show the e ects of diluting H O with CO and CO on the
and m and cm H O bands resp ectively The m band could b e
measured with a simple straight baseline for all cases but for the binary ices the side
bands pro duced were subsequently subtracted using a p olynomial t of order or to
the wing of the m band as a baseline The resultant correction was less than of the
total band intensity in all cases For the m band we used a straight baseline through the
regions cm and cm and for the m band a straight baseline
from to cm was used For the H O CO mixture the CO O C O
b ending mo de at m cm totally obscures the shap e of the long wavelength side
of the H O m band In this case the H O feature was assumed to b e symmetrical
and twice the measured integrated optical depth from its p eak frequency cm
dep ending on temp erature to cm was taken as an estimate of its full integrated
optical depth
For b oth CO and CO the m band is slightly reduced in strength after initial
dep osition at K As the ice is heated however these molecules b egin to di use through
Table Infrared band intensities of CO in binary ices with H O O and CO after
dep osition at K Uncertainties greater than are listed
Ice Mo de A A A
pure
cm molec
H O CO C O stretch
C O stretch
O C O b end
a
combination
a
combination
H O CO C O stretch
C O stretch
O C O b end
combination
combination
b
O CO C O stretch
C O stretch
O C O b end
combination
combination
b
O CO C O stretch
C O stretch
O C O b end
combination
combination
CO CO C O stretch
C O stretch
O C O b end
combination
combination
CO CO C O stretch
C O stretch
O C O b end
combination
combination
b a b
a b a Band is unobservable due to overlap with strong H O band
Listed comp osition is uncertain by a factor of two see text
Figure Measured values of the integrated optical depth of the H O m O H stretching
feature as a function of temp erature in di erent mixtures ratio ed by the integrated optical
depth of the m band in pure H O directly after dep osition at K Solid squares pure
H O empty squares H O CO empty triangles H O CO
and escap e from the H O matrix Schmitt et al Sandford et al and the m
band grows as more H O molecules form hydrogen b onds Just after dep osition at K
the m band is slightly strengthened and the m band is slightly weakened by the new
molecule Both bands approach the pure H O value as they are heated
4. Discussion
4.1 Comparison with previous studies
The results of our exp eriments show that the band strengths of the infrared absorption
features of CO and CO dep end only weakly on the comp osition of the ice mixture in which
they are diluted Overall while the infrared absorption features of molecules may change
considerably in width dep ending on the dilutant molecule Sandford et al Sandford
Allamandola our results show that their band strengths remain quite constant
i e an increase in width is comp ensated by a decrease in depth For example although
Figure As Fig but for the H O m O H b ending mo de
the CO band near cm is more than three times wider in an H O dominated ice
than in pure CO Sandford et al the band strength only varies by a few p ercent It
has also b een shown that the strengths of the and m and cm
bands of H O are only weakly a ected by the presence of CO and CO i e the change is
ab out for all three features for ices where the amount of ap olar sp ecies is comparable
to the amount of H O
The band strengths determined for CO and CO diluted in H O O CO and CO
stay very close to the values for the pure ices This is in contrast with earlier studies
which have shown large increases in the band strengths of these molecules when diluted in
H O i e by for CO and by factors of to for the features of CO Sandford et al
Palumb o Strazulla Sandford Allamandola Since the errors in our
results are in general less than it app ears that these earlier measurements could b e
signi cantly in uenced by the problems involved in the dep osition metho d as discussed in
x It must b e noted that band strength measurements made with these earlier metho ds
for other molecules also indicate strong variations in di erent matrices e g in the case of
CH Hudgins et al On the other hand the band strengths of CO in a CO matrix
do agree well with those measured previously Sandford Allamandola Additional
careful studies should b e p erformed to study band strengths for Other astrophysically
interesting molecules
Figure As Fig but for the H O m libration mo de
It is found that the cm band of CO in an H O CO ice shows a reversible
dep endence on temp erature decreasing by if the temp erature is raised from K
to K Earlier measurements showed a considerably larger decrease of over this
temp erature range Schmitt et al The di erence b etween these results may b e
attributed to the use of sp ecular re ectance for obtaining infrared sp ectra in this earlier
study since it has b een shown that measurements made with this technique give di erent
p eak absorbances of infrared features of ice samples as compared to sp ectra measured in
transmission Kitta Kr atschmer due to interference losses at the interface b etween
the ice sample and the re ecting blo ck surface Hagen et al
When H O is diluted with other molecules the strengths of its absorption bands are
a ected as discussed in x and shown in Figs and We nd that the strength of
the O H stretching band at m in a mixture of H O CO is reduced to of its
value in a pure H O ice and to in an H O CO mixture see Fig The e ects
of dilution on the m H O absorption band have b een studied previously by Greenb erg
et al Using the statistical concentrations of monomers dimers and trimers of
a molecule randomly placed within a simple cubic lattice Behringer Greenb erg
et al have shown that this band will only b e present for mixtures with an H O
concentration ab ove and that the band strength of the m feature will approximately
b e related to the H O concentration f by the semi empirical relation
f f A
mix
A f
pure
where f is the lower limit for pro ducing the m band With f and
for the H O CO and H O CO mixtures Eq yields A A
mix pure
and resp ectively which are close to the measured reductions of and The
consistency of these results may indicate that there is go o d mixing of the sub ject and
dilutant molecules and that there is little di usion taking place up on dep osition b efore the
molecules b ecome xed in the ice lattice
4.2 Astrophysical implications
Our results have several astrophysical implications First the abundance of CO in
the p olar or H O dominated phase of icy grain mantles obtained from the infrared sp ectra
of ices in dense interstellar clouds b ecomes times larger than previously calculated
Toward emb edded sources in dense clouds the amount of CO in p olar ice is then compa
rable to the amount of CO in ap olar ice Tielens et al Chiar et al Also the
CO column densities as calculated by d Hendecourt Jourdain de Muizon toward
three sources b ecome times higher matching or even exceeding that of CO towards the
ob jects studied Next ices toward emb edded sources have b een observed with line of sight
averaged temp eratures up to K Smith et al In this case the band strength
applied to determine the column density of CO in p olar ice should b e taken slightly lower
than the value of cm molec measured at K For example for an average
ice temp erature of K a CO band strength of cm molec is more suitable
Fig Finally in determining H O column densities from the and m interstellar
absorption bands the band strengths for pure H O ice may b e used since the presence of
small abundances of ap olar sp ecies ab out in the ice has little in uence on its band
strengths However the e ects of dilution should b ecome imp ortant if the concentration of
ap olar molecules would b e considerably larger i e Eq as noted by Greenb erg
et al
Ackowledgements We would like to acknowledge fruitful discussions with Peter Jen
niskens This work was partially funded by NASA grant NGR
References
Allamandola L J Sandford S A Valero G J Icarus
Behringer R E J Chem Phys
Breukers R PhD thesis University of Leiden
Chiar J E Adamson A J Kerr T H Whittet D C B ApJ
d Hendecourt L B Allamandola L J A AS
d Hendecourt L B Jourdain de Muizon M A A L
Eiroa C Ho dapp K W A A
Greenb erg J M van b e Bult C E P M Allamandola L J J Phys Chem
Hagen W Tielens A G G M J Chem Phys
Hagen W Allamandola L J Greenb erg J M Ap SS
Hagen W Tielens A G G M Greenb erg J M Chem Phys
Hudgins D M Sandford S A Allamandola L J Tielens A G G M ApJS
Jiang G J Person W B Brown K G J Chem Phys
Kerr T H Adamson A J Whittet D C B MNRAS
Kitta K Kr atschmer W A A
Palumb o M E Strazulla G A A
Sandford S A Allamandola L J ApJ
Sandford S A Allamandola L J Tielens A G G M Valero G J ApJ
Schmitt B Greenb erg J M Grim R ApJ L
Schutte W A Allamandola L J Sandford S A Icarus
Smith R G Sellgren K Tokunaga A T ApJ
Tielens A G G M Tokunaga A T Geballe T R Baas F ApJ
Whittet D C B Dust in the Galactic Environment IOP Publishing Bristol
Whittet D C B Adamson A J Duley W W Geballe T R McFadzean A D
MNRAS
Whittet D C B Bo de M F Longmore A J Baines D W T Evans A Nat
Whittet D C B Longmore A J McFadzean A D MNRAS P
Willner S P Gillet F C Herter T L et al ApJ
Yamada H Person W B J Chem Phys