Hatano Y., Katsumura Y., Mozumder A. (Eds.) Charged Particle and Photon Interactions with Matter - Recent Advances, Applications, and Interfaces

Подождите немного. Документ загружается.

810 Charged Particle and Photon Interactions with Matter

for the nal C

product indicates that the initial reaction of H with C

to form C

is the rate-

controlling process and the following reactions to form the nal C

proceed much faster. This

nding is in accord with the observation of comets Hyakutake and Hale–Bopp; that is, C

and

but no C

were detected in the comae of these comets. Because C

is known to be formed

by the gas-phase reactions, the C

observed in comets is likely to originate from the gas-phase

reaction. In contrast, the saturated hydrocarbons cannot be formed by the gas-phase reactions. The

presence of C

is consistent with the production of C

in icy grain mantles in the natal cloud,

either by the photolysis of CH

-rich ice or by hydrogen-addition reactions to C

condensed from

the gas phase (Gerakines et al., 1996). The model of gas-grain chemistry in dense interstellar clouds

predicts that the grain-surface concentrations of C

and C

are considerably greater than that

of C

(Hasegawa et al., 1992). This result is derived from the assumed larger activation energy

for H + C

(10.0kJ mol

−1

) than that for H + C

(6.2 kJ mol

−1

). Bennett and Mile investigated the

reaction of H with solid C

and C

at 77K by electron spin resonance using a rotating cryostat

(Bennett and Mile, 1973). They found that the ratio of the rate constants k(H + C

)/k(H + C

)

is about 6 × 10

in solid-phase reactions at 77K. The result is in line with ndings that C

was not

detected

as an intermediate product for the reaction of H with C

29.3.2 teMperature dependence on the yield of reaction productS

Figure 29.3 displays the relationship between the yields of C

(%) and the reaction temperature for

reactions H + C

and H + C

. The yield (%) corresponds to the amount of product with respect

to the sample deposited on the substrate. In this experiment, the sample C

or C

was depos-

ited rst at 10K and then H atoms were sprayed over the solid lm after the substrate temperature

was raised from 10K to the reaction temperature. The experiment was not performed above 50K

because

the reactant and C

hydrocarbon products start to sublime above ∼55K.

For reaction H + C

in Figure 29.3, the yield (%) is temperature independent in the range of

50–40K but it shows a sharp increase below 40K. The yield reaches the maximum (38%) at 12K

and decreases below 12K. For reaction H + C

, no reaction product could be detected at 50K.

At40 and 35K, the yields of C

were nearly three orders of magnitude smaller than that at 10K

(24%). With further decrease of temperature, the yield of C

becomes about 0.1% at 30K and

10 20 30

Temperature (K)

Yield of ethane (%)

40 50 60

Figure 29.3 Temperature dependence on the yields of ethane (%) from reactions of H + C

(lled cir-

cles), D+ C

(open circles), H + C

(lled triangles), and D + C

(open triangles). H- and D-atom spray

time, 60min; thickness of the sample lm, 10ML. (From Hiraoka, K. et al., Astrophys. J., 532, 1029, 2000.

Reproduced

with permission of the AAS.)

Chemical Evolution onInterstellar Grains atLow Temperatures 811

shows a steep increase below 25K. The fact that at 50K about 3% of C

was converted to C

for

reaction H + C

, whereas no C

could be detected for reaction H + C

, suggests that rate

constant k(H + C

) may be orders of magnitude larger than k(H + C

) at this temperature. This

in agreement with the nding by Bennett and Mile (1973).

Villa

et al. studied theoretically the tunneling reaction H + C

→ C

(Villa et al., 1998a,b).

They predicted the classical barrier height from 5.0 to 11.3kJ mol

−1

. They pointed out that tunneling

becomes important below ∼500K. Knyazev et al. performed experimental and theoretical studies

ofthe tunneling reaction H + C

(Knyazev and Slagle, 1996; Knyazev et al., 1996). They found

that the tunneling effect for the reaction is evident and recommended the value of the classical bar-

rier for the reaction to be 16.7kJ mol

−1

. The larger value of the classical barrier for reaction H + C

than that for H + C

is in line with the much greater rate for reaction H + C

than for H + C

observed

by Bennett and Mile (1973) and by us (Hiraoka et al., 2000).

The

rate for the formation of C

is a function of the concentration of H on the solid lm

(i.e.,[H]). The results in Figure 29.3 may be interpretable by the increase of the sticking probability

of H atoms at lower temperature. Becker et al. investigated the absorption of H

by condensed solids

of Ar, Kr, Xe, CO

, and C

as a function of temperature (Becker et al., 1972). They found that the

adsorption capacity shows a steep increase starting at about 30–40K and after reaching the plateau

at about 15K, it decreases below ∼15K. Since the polarizability of the H atom (0.67 × 10

−30

) is

smaller than that of the H

molecule (0.8 × 10

−30

), H may have a smaller sticking probability than

. It is highly probable that the temperature dependence of the sticking probability of H atoms

reects

the C

yields in Figure 29.3.

The decrease in the yield of C

below 12K may be mainly due to the decrease of diffusion

rates of H atoms in the solid C

lm. The lack of thermal energy at lower temperatures may

diminish the penetration of H atoms inside the lm. Actually, Bhattacharya et al. (1981) found that

the

trapped H atoms in CH

and CD

matrices become largely immobile below 12K.

29.3.3 reactionS of d atoMS with Solid c

, c

, and c

: iSotope effect

In the low-temperature tunneling reactions, remarkable isotope effects have been found for many

systems (Miyazaki, 1991). In order to obtain information on the isotope effect, the reactions of

Datoms with C

, C

, and C

at 10K were studied (Hiraoka et al., 2000). The D atoms were

produced by introducing D

(99.5%) in the bottleneck discharge tube in Figure 29.1. The quantita-

tive

analyses of the D-atom-substituted products (C

6−n

) were performed by TDS.

When a 10ML thick C

was reacted with D atoms at 10K, C

was formed as a major

product and C

6–n

as minor products with n = 3–6. The yields of C

6−n

are summarized in

Table

29.1 as a function of the D-atom spray time.

The

formation of C

6−n

with n ≥ 3 by the reactions of D atoms with C

(Table 29.1) indicates

that the H atoms of the reactant C

were substituted by the D atoms. If the H-atom abstraction,

H+C

→ C

+ H

, and its analogue take place to some extent, the formation of C

6−n

with

n ≥ 3 can be explained. However, this is less likely because Bennett and Mile (1973) found that the

H-atom addition reaction to the double bonds takes place exclusively and the H-atom abstraction from

was not observed in the reactions of H atoms with olens. Thus, the formation of C

6−n

with

n ≥ 3 can be attributed only to the occurrence of the abstraction reaction of D with ethyl radical, that

is, D + C

D → C

D + HD and their analogues D + C

5−n

→ C

4−n

+ HD with n = 2–4.

In Table 29.1, there is a trend that the relative yields of C

6−n

with larger n increase with

D-atom spray time. The total yield of C

6−n

increases with D-atom spray time and more than

40% of the reactant C

is converted to C

6−n

for a 300min reaction time. At the early stages of

reaction,

the sample surface is mainly composed of C

, and thus D atoms may have no difculty

encountering the reactant C

on the surface to form an ethyl radical. The D atoms on the surface

may be more mobile than inside the lm; consequently, they may have less difculty approaching

the ethyl radical to form ethane by recombination. As the reaction goes on, the surface of the C

812 Charged Particle and Photon Interactions with Matter

lm is being contaminated by the ethane molecules, and D atoms must diffuse inside the lm to

react with ethylene. For the recombination of a D atom with an ethyl radical to occur inside the lm,

the atom and the ethyl radical must approach each other, whereas for abstraction only the H-atom

transfer is required, which may be preferred by tunneling in the solid lm. This explains why the

D-atom

enrichment becomes more favorable with longer reaction time.

Table 29.2, the temperature dependence of the yields of C

6−n

with n ≥ 2, formed by the

reaction of D with 10ML C

lm for 1h, is summarized. In Figure 29.3, the total yields of C

6−n

with n ≥ 2 by the reaction D + C

are shown as a function of reaction temperature. They show a

similar trend as C

from the reaction H + C

, although they are somewhat smaller than those of

, that is, an isotope effect. In Table 29.2, C

is the major product above 30K, and the rela-

tive amounts of C

6−n

with larger n increase rapidly with decrease of temperature. With decrease

of temperature, the sticking probability of D atoms increases, which results in the accumulation of

6−n

(n ≥ 2) products near the top surface of the lm. The slower diffusion of D atoms inside

the lm results in the more preferable occurrence of the abstraction reaction than that of the recom-

bination. This explains the enrichment of D atoms in C

6−n

at a lower temperature in Table 29.2.

table 29.1

d-atom

pray

ime

ependence

on the y

ields

of C

6−n

Formed

from

eaction

of d with C

time (min) yield (%)

relative yield (%)

1 2 91 9 — — —

5 9 79 15 3 2 1

30 25 70 18 7 3 2

60 33 62 22 9 4 3

180 43 61 22 10 4 3

300 44 58 21 9 6 5

Source: From Hiraoka, K. et al., Astrophys. J., 532, 1029, 2000. Reproduced with permission

of the AAS.

Note: Film

thickness, 10

ML;

reaction temperature, 10

table 29.2

temperature

ependence

of the y

ields

of C

6−n

Formed by

reaction

of d with C

temperature (k) yield (%)

relative yield (%)

10 32 61 21 10 5 3

12 36 63 22 8 4 3

15 31 70 19 7 2 2

20 12 80 15 3 1 1

30 3

∼100

— — — —

40 2

∼100

— — — —

48 1

∼100

— — — —

Source: From Hiraoka, K. et al., Astrophys. J., 532, 1029, 2000. Reproduced with per-

mission of the AAS.

Note: Film

thickness, 10

ML;

D-atom spray time, 60

min.

Chemical Evolution onInterstellar Grains atLow Temperatures 813

When the 10ML thick C

was reacted with D atoms, C

was formed as a major product

and C

as minor products. The total yields of C

6−n

with n = 4–6 are summarized

in Table 29.3 as functions of D-atom spray time. As seen in Table 29.3, the yields of C

6−n

from

reaction D + C

are considerably smaller than that of C

from the reaction H + C

. This

indicates that the constant of the rate-controlling reaction, D + C

→ C

D, is smaller than that

of the reaction, H + C

→ C

. Figure 29.3 shows that the yields of C

6−n

from the reaction

D + C

are much smaller than those of C

from H + C

. C

6−n

could not be detected

at reaction temperatures above 20K for the reaction D + C

despite the fact that C

could be

detected up to about 40K for the reaction H + C

. This suggests that the isotope effect in reac-

tions H/D + C

is much more prominent than in the reactions H/D + C

In Table 29.3, the relative yields of C

6−n

(4 ≥ n), with larger n, increase with D-atom spray

time. The enrichment of D content in C

6−n

may be reasonably accounted for by the occurrence

the sequential abstraction reactions of D with vinyl and ethyl radicals.

Table

29.4 summarizes the relative yields of C

6−n

from reaction D + C

as a function of

reaction temperature. The number of D atoms in C

6−n

increases with decrease of reaction tem-

perature.

This trend is similar to the case of reaction D + C

as shown in Table 29.2.

table 29.3

d-atom

pray

ime

ependence

on the y

ields

6−n

Formed from reaction of d with

time (min) yield (%)

relative yield (%)

1 0.4

∼100

— —

15 0.7 84 16 —

60 4 72 25 3

180 12 58 30 12

300 18 55 33 12

Source: From Hiraoka, K. et al., Astrophys. J., 532, 1029,

2000.

Reproduced

with permission of the AAS.

Note: Film

thickness, 10

ML;

reaction temperature, 10

table 29.4

temperature

ependence

the

ields

6−n

Formed

from r

eaction

of d with C

temperature (k) yield (%)

relative yield (%)

10 4 72 25 3

15 1 76 24 3

20 0.1

∼100

— —

∼0

— — —

∼0

— — —

∼0

— — —

Source: From Hiraoka, K. et al., Astrophys. J., 532, 1029, 2000.

Reproduced with permission of the AAS.

Note: Film

thickness, 10

ML;

D-atom spray time, 60

min.

814 Charged Particle and Photon Interactions with Matter

The association reaction, C

+ H

→ C

, may be one of the candidates for the formation

of ethane. However, it has been widely accepted that this reaction is slow (Knyazev and Slagle,

1996; Knyazev et al., 1996), for example, the rate constant of the reaction is of the order of

−16

cm

molecule

−1

even at ∼600K. This is mainly due to the large energy barrier for the reaction

(>33kJmol

−1

) (Knyazev and Slagle, 1996; Knyazev et al., 1996). The contribution of this reaction

may be minor for the formation of ethane.

In order to examine the reactivity of the H atom with C

, the reaction of D atoms with a C

lm was investigated. If reaction D + C

→ C

+ HD takes place, C

D will be formed as a

primary product. When a 10ML thick C

lm was sprayed by D atoms for 1h at 10K, 0.2% of

the reactant C

was converted to C

D. This indicates the occurrence of abstraction reaction,

D + C

→ C

+ HD, but the yield of C

D (0.2%) is much lower than the ethane yields from

the reactions of D with C

(33%) or C

(4%) at 10K for 1h D-atom spray. At temperatures above

20K, no C

D could be detected. The extremely low yield of C

D indicates the much smaller

rate

constant for D + C

than those for D + C

at 10 K.

The negative temperature dependences for the rates of reactions H/D + C

in Figure 29.3

may be reasonably explained either by the increase of the sticking probability of H/D atoms on solid

lms or by the slower rate of the H/D atom diffusion on and in the solid lm at lower temperatures.

It is highly probable that the temperature dependence of the sticking probability and the diffusion

rateof H atoms reect the yields in Figure 29.3. During the H-atom spray over the solid lm, the

number of H atoms sprayed over the lm surface per unit time (d[H]/dt) is equal to the sum of the rates

of the H-atom annihilation due to recombination reaction, H + H → H

, the reaction of H with the

reactant, and the H-atom desorption from the surface. That is, the following relationship should hold:

d H /d constant H H reactant H

[ ] [ ] [ ][ ] [ ]t k k k= = + +

(29.1)

where k

, k

, and k

are the rate constant for reaction H + H → H

, that for the reaction of H withthe

reactant, and that for the desorption of H from the surface, respectively. The experimental fact that

the rates of the tunneling reactions increase with decrease of temperature (Figure 29.3) means that

the value of k

[H][reactant] in Equation 29.1 increases with decrease of temperature. That is, k

and/or [H] increase with decrease of temperature. In Equation 29.1, k

must be highly temperature

dependent, that is, k

= A exp(−E

/RT) (A, pre-exponential factor; E

, activation energy for the

desorption of the H atoms from the surface). At high temperature, the larger k

keeps the steady-

state concentration of H ([H]) on the solid surface extremely low. With decrease of temperature, k

decreases exponentially, and concomitantly [H] on the sample surface increases. Besides, the slower

diffusion rate of H atoms on and in the lm at lower temperatures results in the smaller value of k

in Equation 29.1. This also leads to the increase of the H-atom concentration at lower temperatures.

In Figure 29.3, the yield of C

from reaction H + C

is of the same order as that from H +

at 10K. In case k

for H + C

is much larger than k

for H + C

, as predicted by Bennett and

Mile (1973), [H] on the C

lm must be much larger than [H] on the C

lm at ∼10K. In order to

examine this possibility, the reactions of H atoms with a few ML thick C

or C

deposited on the

were studied (Hiraoka and Sato, 2001).

If the sticking probability of H atoms on the C

is much higher than that on the C

lm, the

conversion of C

to C

for the lm C

would become much larger than that for C

. The yields of C

were 22% and 28% for the C

(3ML)/C

(10ML) and C

(3ML)/C

(10ML) lms, respectively. This suggests that the steady-state concentrations [H] on the C

and

lms are of the same order. Thus, the steep increase of the yield of C

from reaction H + C

could possibly be due to the negative temperature-dependent rate constant k

for H + C

. A nega-

tive temperature dependence of the rate constants for the low-temperature tunneling reactions has

been predicted theoretically by Takayanagi and Sato (1990). They performed the bending-corrected

rotating-linear-model calculations of the rate constants for the H + H

reaction and its isotopic vari-

ants at low temperatures and examined the effect of the van der Waals well. They found that van der

Chemical Evolution onInterstellar Grains atLow Temperatures 815

Waals wells included in both potential surfaces signicantly affected the calculated rate constants at

temperatures lower than 10K.

29.4 Formation oF Formaldehyde by the tunneling

Ctions

oF h with s

olid

at 10 k

Formaldehyde (H

CO) is one of the most complex molecules for which specic gas-phase reaction

+ O → H

CO + H has been proposed (Watson, 1977). The production of H

CO is then crucially

dependent on the adequate production of CH

. On the other hand, the signicant abundances of

CO in the diffuse envelopes of dark clouds cannot be explained by purely gas-phase processes.

The production of CH

is insufcient in these regions to drive the reaction to form enough H

(Sen

et al., 1992).

After

, CO is the most abundant molecule in dense clouds and thus has special importance

(Tielens et al., 1991). Tielens et al. (1991) suggested that most of the CO accreted in H

O-rich

mantles has reacted with other species on the grain surface. When the H-atom accretion rate is high,

this leads to the formation of HCO, H

CO, and possibly CH

OH, accounting for the observed large

abundance

of CH

OH in grain mantles (Tielens, 1989) and comets (Crovisier, 1998).

CO HCO H CO H CO CH OH

H H H H

 

 

 

 

 

 

 

 

2 3 3

(29.2)

Van Ijzendoorn et al. (1983) measured the absorption spectrum of HCO in Ar, Kr, Xe, CH

, CO, and

matrixes. They found that the absorbance of HCO in Ar, Kr, and CH

matrices grew several-fold

while

that for trapped H atoms decreased on warm up from 10 to 15

In Sections 29.4.1 through 29.4.3, it will be shown that the low-temperature tunneling reaction of

Hwith CO may not be the major source for the ubiquitous interstellar methanol (Hiraoka et al., 2002a).

29.4.1 forMation of h

co froM the reaction of h with Solid co

It should be noted that the CO molecule is highly adsorptive (Hiraoka et al., 2002a). The inner wall

of the H-atom spray bottleneck discharge was found to be rather quickly contaminated by the CO

molecules. With the contamination of the discharge tube, several hydrogenated products such as

and CH

OH were formed when the H

plasma was generated in the bottleneck discharge tube.

The gaseous products formed in the H

plasma were sprayed over the silicon substrate and con-

densed there as contaminants. Thus, great care must be taken for the measurement of the reaction of

H with CO. Whenever the sample gas was deposited on the silicon substrate, the exit of the H-atom

spray bottleneck discharge tube was plugged by the at stainless steel lid that prevented the sample

gas molecules from entering into the discharge tube (see Figure 29.1). By using this experimental

system, no discharge products were detected for a long period of the repetitive experiments.

Figure 29.4 shows the TDS spectra of the 10ML thick solid CO sample reacted with H atoms for

1h at 10K. The peak at m/z 30 appearing at ∼130K is due to the formation of H

CO. If methanol is

formed as the reaction product, the base peak at m/z 31 for methanol must appear in the desorption

mass spectrum. However, no ion signals with m/z 31 could be detected by TDS as shown in Figure 29.4.

This indicates that the reaction of H atom with H

CO to form CH

OH is negligible under the present

experimental conditions.

We investigated the reaction of H atoms with C atoms seeded in the CO solid (Hiraoka et al.,

1998). It was found that the H atoms diffuse deep inside the CO solid lm and hydrogenate the

Catoms embedded in the CO matrix efciently. This clearly indicates that the H atoms can migrate

in the CO matrix and have ample chance to react with seeded C atoms to form CH

. In other words,

the

reactivity of H atom toward the CO molecule is low enough to allow the diffusion of H atom in

the

CO matrix.

816 Charged Particle and Photon Interactions with Matter

29.4.2 teMperature dependence of the yield of h

co froM the reaction h + co

Figure 29.5 displays the relationship between the yields of H

CO (%) and the reaction temperature

for reaction H + CO together with the yields of C

, C

, and solid product (mainly polysilane) for

reactions of H + C

, H + C

(Hiraoka et al., 2000), and H + SiH

(Hiraoka et al., 2001a), respec-

tively, obtained under similar experimental conditions. The experiment above 25 K for H+CO was

not made because the reactant CO starts to sublime above ∼28K. The yield of H

CO in Figure29.5

shows a steep increase with decrease of temperature from 0.01% at 25K to 0.08% at 10K. In

Figure29.5, one sees a general trend that the rates of all tunneling reactions dealt with increase with

a decrease of temperature. In general, a steeper increase of the rates with a decrease of temperature

was observed for molecules whose rates of reactions with H are small at higher temperature regions,

for example, C

and CO in Figure 29.5. The observed increase of the rates of reactions H + C

and H + CO in Figure 29.5 is much steeper than the increase of the sticking probability of H atoms

on amorphous carbon with a decrease of temperature (Pirronello et al., 2000), that is, the observed

increases of the yields of C

and H

CO in Figure 29.5 may not be explainable only by the increase

the sticking probability of the H atoms on the solid surface.

29.4.3 doeS the van der waalS Solid filM erode during the h-atoM Spray?

In the low-temperature reactions, only exothermic reactions can take place. A greater part of the

heat of reaction evolved will eventually degrade to the phonon energy. There is ample evidence that

multiphonon processes for the dissipation of the vibrational energy are slow because of the large

differences in the vibrational and cohesive energies (Dressler et al., 1975; Oehler et al., 1977).

Calaway and Ewing (1975) found that the lifetime of the decay of v = 1−0 vibrational levels of N

is as long as 1.5 ± 0.5s in liquid nitrogen. In the present experiment, the H-atom recombination

reaction H + H → H

with heat of reaction of 435kJ mol

−1

, as well as other exothermic reactions

such as reactions (29.2), take place on the solid surface. The occurrence of exothermic reactions on

the solid surface may result in the local heating of the van der Waals solid. This could lead to the

erosion of the sample lm. Actually, this process is considered to be one possible mechanism for the

desorption

of adsorbed molecules into the gas phase in the dark clouds at 10

0.0

5.0×10

–10

1.0×10

–9

1.5×10

–9

2.0×10

–9

1×10

–12

2×10

–12

3×10

–12

4×10

–12

m/z 28

m/z 30

m/z 31

×500

100

Temperature (K)

Pressure (Torr)

150 200

Figure 29.4 TDS spectra for the 10M thick CO solid sample reacted with H atoms for 1 h at 10 K.

(From Hiraoka,

K. et al., Astrophys. J. 577, 265, 2002a. Reproduced with permission of the AAS.)

Chemical Evolution onInterstellar Grains atLow Temperatures 817

In order to check whether the desorption of CO molecules takes place during the H-atom spray

over the solid CO lm, the recovery of deposited CO was measured after the solid CO lm was

sprayed by the H atoms for 1h by means of TDS. Figure 29.6 shows the temperature dependence of

the ratio of the CO recovered after the H-atom spray for 1h at a certain temperature to that for the

deposited at 10K without spraying the H atoms. As seen in Figure 29.6, the ratio is almost indepen-

dent of the temperature up to 25K. This suggests that the relaxation of the vibrational energy of

formed by the recombination reaction

H H H

+ →

to the lattice phonon is inefcient. This nding is

in accord with the theoretical calculation by Takahashi et al. (1999). They investigated the time and

space dependence of the local temperature increase of icy mantles caused by the release of H

for-

mation energy in the vicinity of H

-forming sites using classical molecular dynamics computational

simulations. They predicted that the H

formation energy was partitioned to the vibrational energy

(78.5%), the rotational energy (9.9%), and the translational energy (7.4%) for the desorbing H

mol-

ecule and only 4.4% was absorbed to the amorphous ice. The contribution of the highly exothermic

reaction H + H → H

to the desorption of the grain mantle may not be as large as thought before.

29.5 reaCtion oF h and d atoms with solid Formaldehyde

and

ethanol

at Cryogeni

C temperatures

In Section 29.4, studies on the reaction of H with solid CO at 10K (Hiraoka et al., 2002a) have

been described. The CO molecules were found to be converted to H

CO with a very low efciency.

Besides,

no CH

OH was detected within the experimental error.

In this section, the study on the reactions of H and D atoms with solid H

CO and CH

OH lms

will be discussed (Hiraoka et al., 2005). It would give us more information on the reaction mecha-

nisms for the formation of H

CO and CH

OH from reaction H + CO in the dark clouds because

CO is the intermediate product for the reaction of H with CO to form CH

OH.

10 20 30

Temperature (K)

SiH

Yield (%)

40 50 60

0.10

0.08

0.06

0.04

0.02

0.00

Yield of H

CO (%)

Figure 29.5 Relationship between the yield of H

CO (%) and the reaction temperature for reaction H + CO.

The yields of C

, C

, and solid product (mainly polysilane) for reactions H + C

, H + C

(Hiraoka

et al., 2000) and H + SiH

(Hiraoka et al., 2001a), respectively, are also shown. These experiments were made

under similar experimental conditions. Film thickness, 10ML; H-atom spray time, 1h. (From Hiraoka, K. et al.,

Astrophys. J.,

577, 265, 2002a. Reproduced with permission of the AAS.)

818 Charged Particle and Photon Interactions with Matter

The sample gas H

CO was prepared by heating the paraformaldehyde powder (Aldrich, 95%) in

a glass bulb at about 70°C–100°C. The moisture in the formaldehyde vapor was removed by putting

some unhydrous magnesium sulfate together with the paraformaldehyde powder in the glass bulb.

The

formaldehyde or methanol sample was degassed by repeating the freeze-pump-thaw cycles.

The

product analysis was performed by TDS. For the H

CO sample, less-volatile paraformalde-

hyde was found to be the major product by FT-IR. The yield of paraformaldehyde was calculated by

comparing

the TDS spectra of the neat sample (not reacted) with that being irradiated by H atoms.

29.5.1 reaction of h with Solid h

Figure 29.7a shows the FT-IR spectra of 20 ML thick H

CO sample sprayed by H atoms for 300min

at 10 K. The bottom spectrum corresponds to that for the 20ML thick H

CO before the H-atom

spray. With increase in reaction time, the intensities of H

CO absorption (1182 and 1246cm

−1

)

decrease gradually and the growth of new absorption for CH

OH (1035cm

−1

), paraformaldehyde

(1126cm

−1

), and CO (2140cm

−1

) is observed.

The chemical reactions envisaged to occur are summarized in Table 29.5. The formation of

methanol suggests the occurrence of reactions (29.3) through (29.5) in Table 29.5. Reaction (29.3b)

is more exothermic than reaction (29.3a) by about 41kJ mol

−1

because the bond energy of H–CH

(393kJ mol

−1

) is smaller than that of CH

O–H (435kJ mol

−1

). However, Sosa and Schlegel (1986)

theoretically predicted that the barrier height for H + H

CO → CH

O (19.2kJ mol

−1

) is consider-

ably lower than that for H + H

CO → CH

OH (43.9kJ mol

−1

) at the MP4SDTQ/6-31(d) after spin

annihilation. Woon (2002) performed an ab initio calculation and predicted that reaction (29.3b) has

a barrier about 42kJ mol

−1

higher than reaction (29.3a) that excludes reaction (29.3b). At present, no

information

is available on the branching ratio of reaction (29.3a) and (29.3b).

The

appearance of CO in Figure 29.7a suggests the occurrence of H-atom abstraction reactions

(29.6) followed by reactions (29.7) and/or (29.8a). Here, reaction (29.7) is the H-atom abstraction reac-

tion while reaction (29.8a) is the recombination reaction of two radicals, H and HCO, to form the

0.0

0.2

0.4

0.6

0.8

1.0

1.2

5 10 15

Temperature (K)

Ratio

20 25 30

Figure 29.6 Temperature dependence of the ratio of the CO recovered after the reaction of CO lm with H

atoms for 1 h at a certain reaction temperature to that for the CO deposited at 10K without spraying the Hatoms.

(From Hiraoka, K. et al., Astrophys. J., 577, 265, 2002a. Reproduced with permission of the AAS.)

Chemical Evolution onInterstellar Grains atLow Temperatures 819

intermediate activated complex [H

CO]* followed by its unimolecular dissociation. In reaction (29.9),

the intermediate complex [H

CO]* is stabilized through the excess energy dissipation to the sur-

rounding matrix molecules. It should be noted that the reaction H

CO → H

+ CO is exothermic

by 2kJmol

−1

, that is, the decomposition of H

CO into H

and CO is thermochemically favorable.

However, the molecular fragmentation channel (29.8a) is found to be minor relative to the atomic

fragmentation channel (29.8b) in the photoexcitation fragmentation of H

CO (Fleck et al., 1991). The

branching ratio of reactions (29.7) through (29.8a) could not be determined in the present experiment.

2200

0.000

0.002

0.004

0.006

0.008

0.010

2100

2140

1246

1182

(CH

1126

1035

(CH

951

300 min

180 min

120 min

60 min

30 min

10 min

H spray

Before reaction

1300 1200

Wavenumbers (cm

–1

)(a)

Absorbance

1100 1000 900

2200

0.000

0.002

0.004

0.006

0.008

0.010

2100

2140

1246

1182

(CH

1126

1035

(CH

951

(CH

1232

230 K

210 K

190 K

170 K

150 K

130 K

110 K

90 K

70 K

50 K

30 K

10 K

After H spray

1300 1200

Wavenumbers (cm

–1

)(b)

Absorbance

1100 1000 900

Figure 29.7 (a) FT-IR spectra for a 20ML thick H

CO lm sprayed by H atoms for 300 min at 10K. The

bottom spectrum corresponds to that for the 20 ML thick H

CO lm before the reaction. (b) Temperature

dependence of the FT-IR spectra for the 20ML thick H

CO lm after the lm was sprayed with H atoms for

300min at 10 K. The substrate temperature was increased continuously from 10 to 230K with a heating rate of

6K min

−1

. The rapid growth of peaks at ∼951 and ∼1126cm

−1

observed at ∼70K is due to the radical-induced

polymerization

reaction of H

CO to form paraformaldehyde.

(continued)