VUV photoionization time-of-flight mass spectrometry of flash pyrolysis of silane and disilane

Article abstract of DOI:10.1016/S0009-2614(01)00739-4

Flash pyrolysis of silane, SiH₄, and disilane, Si₂H₆, diluted in He or Ar (1%), was carried out at temperatures ranging from ~700 to ~1500 K. After a short reaction time of ~20 μs, the initial products were isolated in a supersonic molecular beam and detected by single vacuum ultra-violet (VUV) photon (λ=118or121nm) ionization time-of-flight mass spectrometry (TOFMS). Initial decomposition and reaction products, both free radical intermediates and stable species, were directly observed, which included SiH₂ and Si₂H₄.

Full text of DOI:10.1016/S0009-2614(01)00739-4

10 August 2001
Chemical Physics Letters 343 /2001) 482±488
www.elsevier.com/locate/cplett
VUV photoionization time-of-¯ight mass spectrometry
of ¯ash pyrolysis of silane and disilane
Steven D. Chambreau, Jingsong Zhang *,1
Department of Chemistry, University of California, Riverside, CA 92521, USA
Received 3 April 2001; in ®nal form 23 May 2001
Abstract
Flash pyrolysis of silane, SiH4, and disilane, Si2H6, diluted in He or Ar /1%), was carried out at temperatures ranging
from ꢀ700 to ꢀ1500 K. After a short reaction time of ꢀ20 ls, the initial products were isolated in a supersonic mo-
lecular beam and detected by single vacuum ultra-violet /VUV) photon /k ˆ 118 or 121 nm) ionization time-of-¯ight
mass spectrometry /TOFMS). Initial decomposition and reaction products, both free radical intermediates and stable
species, were directly observed, which included SiH2 and Si2H4. Ó 2001 Elsevier Science B.V. All rights reserved.
1. Introduction
plications present in mechanistic studies in the
pyrolysis experiments include recombination of
reactive species before their detection, and frag-
mentation of species upon ionization by high-en-
ergy electron impact ionization /EI, typically 70
eV) or by multi-photon ionization /MPI) tech-
niques. Fragmentation of ionic species complicates
the mass spectra and makes it dicult to distin-
guish contributions from pyrolysis products versus
from ion fragmentation.
The experimental approach in this work mini-
mizes e€ects of recombination by using a short
reaction time of approximately 20 ls [19±21]. The
initial pyrolysis products are subsequently isolated
by `freezing' them upon supersonic expansion into
a molecular beam /with T < 50 K). In addition to
the cooling of internal energy, ion fragmentation is
further diminished by utilizing a `soft' vacuum
ultra-violet /VUV) photoionization source, with
photon energies of 10.2 and 10.45 eV. These en-
ergies are sucient to ionize almost all of the sil-
icon hydrides [both free radicals and stable species,
SinHx ꢁn ˆ 1±2; x ˆ 0±6†, except SiH4] with little
Reactions involving silicon hydrides are of
considerable importance in processes such as
epitaxial growth of Si thin ®lms, formation of Si
nanoparticles, and growth of amorphous Si ®lms
[1]. Applications of these processes include man-
ufacture of microelectronics, solar cells, and opti-
cal sensors and displays. Understanding of the
mechanisms involved in these processes is critical
in improving control over the desired properties of
the products.
Homogeneous gas-phase thermal decomposi-
tion of silicon hydrides is an important pathway
involved in Si growth processes. Much theoretical
and experimental work has been done on both
disilane [2±11] and silane [12±18] pyrolysis. Com-
* Corresponding author. Fax: +1-909-787-4713.
E-mail address: jszhang@ucrac1.ucr.edu /J. Zhang).
1
Also at Air Pollution Research Center, University of
California at Riverside, Riverside, CA 92521, USA.
0009-2614/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 9 - 2 6 1 4 / 0 1 ) 0 0 7 3 9 - 4

S.D. Chambreau, J. Zhang / Chemical Physics Letters 343 +2001) 482±488
483
or no fragmentation [22±25]. Thus, this experiment
allows direct observation of the initial pyrolysis
products, especially free radical intermediates
[19±21].
In this Letter, we report the results of ¯ash
pyrolysis of silane and disilane using single VUV
photon ionization time-of-¯ight mass spectrome-
try /TOFMS). Reactive intermediates and pyrol-
ysis products, such as SiH2 and Si2H4, have been
directly observed.
O.D., 1 mm I.D., Carborundum). The initial
products and undecomposed parent molecules
exited the heated tube and supersonically ex-
panded into the vacuum chamber, where they were
skimmed and passed into the photoionization re-
gion of the TOFMS /Fig. 1). The pyrolysis prod-
ucts were ionized by VUV radiation at k ˆ 118 nm
/10.45 eV) or 121 nm /10.2 eV). The 118-nm ra-
diation was produced by frequency tripling the
355-nm output from a Nd:YAG laser in a Xe cell
/ꢀ50 Torr), and the 121-nm by tripling 365-nm
radiation from a dye laser in a Kr cell /ꢀ80 Torr).
The VUV radiation was focused by a MgF2 lens
through a small aperture into the photoionization
zone, while the fundamental 355-nm or 365-nm
UV beam diverged in this region. This divergence
and the aperture minimized MPI and the amount
of UV radiation and its scattering in the ionization
zone. Without Xe or Kr, the tripling medium for
VUV generation, essentially no ion signals were
observed for SiH4 and Si2H6 at all temperatures,
indicating negligible MPI and secondary EI by the
fundamental UV radiation alone. MPI and frag-
mentation due to both VUV and UV photons /e.g.,
2. Experimental
The pyrolysis experiments were performed with
two linear TOF mass spectrometers with two dif-
ferent
VUV
photoionization
wavelengths
/k ˆ 118 nm: TOFMS, Model D-651, the R. M.
Jordan Company [26]; k ˆ 121 nm: home built
TOFMS), and have been described previously
[27,28]. TOF mass spectra were recorded on a
digital storage oscilloscope /Tektronix TDS3032,
300 MHz) or a multi-channel scaler /EG&G,Tur-
bo-MCS), and were averaged from 500 to 2000
laser shots.
‡
‡
UV photodissociation of SiH4 and Si2H6 parent
ions after VUV photoionization) was also negli-
gible. This was supported, for example, by little or
no fragmentation in the room-temperature Si2H6
Flash pyrolysis was carried out by expanding
the silicon hydride precursors, seeded in either He
or Ar /1%), through a heated SiC tube /2 mm
Fig. 1. Flash pyrolysis/supersonic jet/VUV photoionization TOFMS apparatus.

484
S.D. Chambreau, J. Zhang / Chemical Physics Letters 343 +2001) 482±488
‡
‡
‡
mass spectra at 121 and 118 nm /Figs. 2 and 3). On
the contrary, if the VUV + UV MPI fragmentation
prevailed, the mass patterns expected at the
VUV + UV photon energy /13.6 eV for 121 + 365
nm and 14.0 eV for 118 + 355 nm, with a maxi-
mum of 0.5 or 0.7 eV taken away by a photo-
Si2H2 :Si2H4 :Si2H6 ꢀ 4:3.7:1) [23]. This observa-
tion of negligible VUV ‡ UV MPI and fragmen-
tation was consistent with those in previous studies
[11,21]. The design of the pyrolysis nozzle was
based on that of Chen and co-workers [20]. The
temperature was monitored with a type c ther-
mocouple /Omega) which was calibrated to the
SiC tube's internal temperature. Assuming near
sonic ¯ow velocity within the heated region, the
residence time of the parent molecules was esti-
mated to be ꢀ 20 ls [20]. After performing several
series of silane and disilane pyrolysis experiments,
the inside of the SiC tube would become coated
with amorphous Si along the heated section. In
order to ensure that the Si deposit did not a€ect
our results, the data presented in this Letter were
obtained with new SiC tubing and before signi®-
cant Si buildup in the tube.
‡
electron, respectively) would be mainly the Si2H2
‡
and Si2H4 fragments /with
a
ratio of
The silane /99.998%, Matheson) and disilane
/99.998%, Aldrich Chemicals) gas samples were
diluted to 1% mixtures in helium or argon without
further puri®cation. The gas mixture was intro-
duced into the pyrolysis nozzle with a backing
pressure of ꢀ1.3 atm, and the vacuum chamber
was held at less than 5 Â 10À4 Torr with the mo-
lecular beam on.
Fig. 2. Mass spectra of pyrolysis of disilane /1% in He) be-
tween room temperature and 1120 K using 10.2 eV photoion-
ization. The mass spectra have been o€set in both mass and
baseline for clarity.
3. Results
3.1. 10.2 eV Photoionization mass spectrum of Si2H6
pyrolysis at 300±1120 K
As shown in Fig. 2, when the heater is at room
temperature /300 K), the observed peaks
m=e ˆ 62; 63, and 64 are due to the Si2H6 pre-
cursor /1% in He). The intensities of these peaks
agree with the Si isotopic abundances /92.2%,
4.7%, and 3.1% for 28Si, 29Si, and 30Si, respec-
tively). A H-atom peak /m=e ˆ 1) is mainly from
background gas /con®rmed with the molecular
beam o€) and changes little with the heater tem-
peratures; its high intensity is due to the sensitive
resonantly enhanced MPI of H atoms [121.6 nm
/1s ! 2p) + 365 nm]. As discussed earlier and
similar to a previous study [11], absence of
photoionization fragments of Si2H6 in the TOF
mass spectrum at 300 K indicates negligible
Fig. 3. Mass spectra of pyrolysis of disilane /1% in Ar) between
room temperature and 1440 K using 10.45 eV photoionization.
The spectra have been o€set in both mass and baseline for
clarity.

S.D. Chambreau, J. Zhang / Chemical Physics Letters 343 +2001) 482±488
485
contribution from VUV ‡ UV MPI. In addition,
Si2H6 and its thermal dissociation products at el-
evated temperatures can be eciently cooled by
supersonic expansion, minimizing photoionization
fragmentation of these species.
[23,29], the observed small intensities of m=e 30,
31, and 58±60 peaks relative to that of m=e 62 at
300 K indicate minor contributions of EI and/or
VUV + UV MPI in our TOF mass spectra.
Due to the supersonic cooling and low tem-
perature of the molecular beam, the extent of EI of
Si2H6 and its fragmentation, as well as MPI frag-
mentation /if any), should remain small and nearly
constant at elevated pyrolysis temperatures, as
indicated, for example, by the unvarying low in-
tensities of m=e 30 and 31 up to T ˆ 1065 K. Thus,
appearance and growth of mass peaks result from
production of neutral species in the pyrolysis,
which are subsequently photoionized. The m=e 62
peak remains nearly constant up to 935 K, and
declines rapidly up to 1440 K to less than 5% of
the peak height at 935 K. The m=e 60 peak grows
from 935 K, and then declines rapidly above 1200
K. The m=e 58 peak begins to grow at 935 K,
which, similar to m=e 60, increases up to
T ˆ 1200 K, and declines up to 1440 K. At 1065
K, a m=e 56 peak appears, which grows to 1200 K,
and then declines slightly at higher temperatures.
At T > 1065 K, a m=e 30 peak starts to grow, and
declines slightly at 1440 K. Within the same tem-
perature range /1200±1440 K), a peak appears at
m=e 28, whose height does not change signi®cantly
as the temperature is increased. These results
con®rm the production of SiH2; Si2H4, and Si2H2
upon pyrolysis of Si2H6. Above 1200 K, the ap-
pearance of mass 28 and 56 peaks signi®es pro-
duction of Si and Si2, respectively. Comparison of
these results to the 10.2 eV photoionization spectra
gives good agreement and leads to similar impli-
cations concerning mechanisms involved in disi-
lane pyrolysis.
As the pyrolysis temperature is raised to 725 K,
an additional peak appears at m=e 60, corre-
sponding to Si2H4. The m=e 60 peak increases with
temperature up to 925 K /with a small isotopic
satellite peak at m=e 61), and then declines above
990 K. The growth of an m=e 58 peak occurs at
ꢀ800 K, indicating the formation of Si2H2. Also,
at 1055 K, there appears a peak at m=e 56 which is
assigned to Si2. At 925 K, a peak at m=e 30 ap-
pears, corresponding to production of the SiH2
radical, and this peak reaches a maximum at 1055
K. Correspondingly, the parent m=e 62 peak de-
clines rapidly above 855 K due to the ecient
pyrolysis.
3.2. 10.45 eV Photoionization mass spectrum of
Si2H6 pyrolysis at 300±1440 K
10.45 eV Photoionization mass spectrum of di-
silane in Ar /1%) with the heater at room tem-
perature shows peaks at m=e 62, 63, and 64 /from
Si2H6), and in addition, small peaks at 20, 30, 31,
40, 58, 59, and 60 /Fig. 3). The m=e 40 and 20
‡
peaks indicate ionization of Ar /Ar and Ar2‡),
which has a ®rst ionization potential /IP) of 15.8
eV [22], and could not be caused by 10.45 eV
photoionization. These peaks are attributed to EI
of Ar due to a small amount of photoelectrons
caused by scattered VUV radiation within the
photoionization region. Their broader widths are
consistent with the less well-de®ned EI volume, as
opposed to the VUV photoionization. Similarly,
‡
‡
since the appearance energy /AE) of SiH2
‡
3.3. 10.45 eV Photoionization mass spectrum of
SiH4 pyrolysis at 300±1495 K
/m=e ˆ 30),
SiH3
/m=e ˆ 31),
HSiSiH
‡
/m=e ˆ 58), and H3SiSiH /m=e ˆ 60) from Si2H6
is 12.0, 11.8, 11.5, and 10.8 eV, respectively [22],
these minor peaks should be produced from EI
and/or MPI fragmentation of Si2H6. The AE of
H2SiSiH2 from Si2H6 is 10.04 eV, but its disso-
ciative ionization cross-section is very small [23],
Photoionization mass spectra at 10.45 eV of
SiH4 pyrolysis are shown in Fig. 4. At 300 K,
peaks at m=e ˆ 28±31 are the result of EI of SiH4,
‡
‡
‡
‡
‡
corresponding to Si ; SiH ; SiH2 , and SiH3
fragments, respectively. Between 300 and 830 K,
the intensities of these peaks remain nearly con-
stant. At 935 K, the m=e 30 peak starts to grow,
and above 1200 K, it increases signi®cantly, clearly
consistent with the negligible m=e 60 peak at 300
K. Since Si2H4 and Si2H2 are the main fragments
‡
‡
of Si2H6 in ꢀ100 eV EI or ꢀ13 eV VUV + UV MPI

486
S.D. Chambreau, J. Zhang / Chemical Physics Letters 343 +2001) 482±488
The de®nitive proof of this channel is the ap-
pearance of the m=e 30 peak in the 10.2 eV
photoionization mass spectra /Fig. 2). Under these
conditions, EI and MPI are absent, and photoi-
onization fragmentation is minimal. To our
knowledge, this is the ®rst direct mass spectro-
metric observation of the SiH2 radical in the py-
rolysis of disilane. Due to its high IP of 11.0 eV
[22,25], the SiH4 product is not detected.
At the photoionization energy of 10.45 eV, the
increase in relative peak ratios of mass 30 versus
mass 62 with increasing temperature is also a
strong indication of SiH2 production in Si2H6
‡
pyrolysis. As discussed previously, the small SiH2
‡
Fig. 4. Mass spectra of pyrolysis of silane /1% in He) between
room temperature and 1495 K using 10.45 eV photoionization.
The spectra have been o€set in both mass and baseline for
clarity.
and SiH3 signals at temperatures less than ꢀ1000
K are attributed to a minor amount of EI or MPI.
Although signi®cant internal energy is imparted to
disilane upon heating, disilane should be cooled by
supersonic expansion and the EI and photoion-
ization fragmentation patterns should not change
signi®cantly with increasing temperature, as indi-
cated by the nearly constant relative ratios of m=e
30, 31, and 62 up to ꢀ930 K. Thus, the increase of
m=e 30 peak at temperatures above 1000 K, in
contrast to the signi®cant decrease of m=e 31 and
62 peaks, is not due to fragmentation of internally
hot disilane ions, but rather from the pyrolysis of
Si2H6.
indicating the production of SiH2. Above 1335 K,
the m=e 31 peak also increases, which is due to
28SiH3. At 300 K, an m=e 62 peak is possibly due
to a small amount of disilane impurity, or more
likely, it is attributed to EI or photoionization
fragmentation of SiH4 clusters in the molecular
‡
‡
beam, e.g., ꢁSiH4† ‡ hm=eÀ ! ‰ꢁSiH4† Š ! Si2H6
2
2
‡H2. The latter possibility is supported by the
decrease of the m=e 62 peak, presumably following
the decreased SiH4 cluster formation in the beam,
as the temperature is raised to 715 K. The m=e 62
peak grows from 830 to 935 K and then declines
above 1070 K, implying both production and py-
rolysis of Si2H6. Peaks at m=e 58 and 60 appear
and increase above 830 K, corresponding to Si2H2
and Si2H4 products. In addition, m=e 56 peak
grows above 1070 K, indicating Si2 product.
4.2. Dehydrogenation products of Si2H6: Si2H4,
Si2H2, and Si2
At 10.2 eV, the appearance of mass 60 at ele-
vated temperatures indicates the production of
Si2H4, another important pathway of Si2H6 py-
rolysis [2±11]:
Si2H6 ! H3SiSiH ‡ H2;
H3SiSiH ! H2SiSiH2:
ꢁ2†
ꢁ3†
4. Discussion
At 10.45 eV, using an argument similar to that for
the production of SiH2, the increase in relative
peak height ratios of mass 60 versus mass 62 with
increasing temperature also indicates the produc-
tion of Si2H4. Although it is not possible to dis-
tinguish between the production of H3SiSiH or
H2SiSiH2; H3SiSiH is likely to be produced ®rst
via a lower activation barrier, and it then can
4.1. SiH2 production from Si2H6
Production of SiH2 radical is believed to be an
important initial pathway in pyrolysis of disilane
[2±11]:
Si2H6 ! SiH4 ‡ SiH2:
ꢁ1†

S.D. Chambreau, J. Zhang / Chemical Physics Letters 343 +2001) 482±488
487
readily isomerize to the energetically favored
H2SiSiH2 [2,11]. H2 is undetectable due to its high
IP of 15.4 eV [22]. Our results con®rm the two
accepted initial thermal decomposition pathways
of disilane: channels /1)±/3) [2±11]. Note that
channel /1) has a lower activation energy [2±11],
thus the appearance of SiH2 at higher tempera-
tures than Si2H4 might be due to a lower detection
sensitivity of SiH2 by the VUV photoionization.
As shown in Figs. 2 and 3, subsequent loss of H2
leads to the formation of Si2H2 and Si2:
Finally, formation of polysilanes /Si3H8; Si4H10,
etc.) and small Sin and SinHx ꢁx < 2n ‡ 2† clusters
is also observed in the pyrolysis of both disilane
and silane, presumably from SiH2 insertion into
the Si±H bond of polysilanes and subsequent de-
hydrogenation [10±18]. This will be discussed in a
future publication.
5. Conclusion
Using a ¯ash pyrolysis source, coupled with
VUV photoionization TOFMS technique, direct
observation of reactive intermediates and initial
products formed upon dissociation of silane and
disilane has been achieved. These results demon-
strate the formation of SiH2, as well as
Si2H4; Si2H2, and Si2, in the pyrolysis of both si-
lane and disilane. This study provides direct evi-
dence in support of previously proposed pathways
involved in the decomposition of silicon hydrides,
namely SiH4 ! SiH2 ‡ H2; Si2H6 ! SiH4 ‡SiH2;
SiH2 ‡ SiH4 ! Si2H6, etc.
Si2H4 ! Si2H2 ‡ H2;
Si2H2 ! Si2 ‡ H2:
A detailed discussion of these mechanisms will be
provided in a future publication.
ꢁ4†
ꢁ5†
4.3. Pyrolysis of SiH4
In Fig. 4, the m=e 28±30 peaks at 300 K are
‡
‡
‡
attributed to the Si ; SiH , and SiH2 fragments
upon EI of SiH4, and these peaks remain at almost
the same intensities up to T ꢀ 830 K. Above ꢀ935
K, the m=e 30 peak starts to grow rapidly, while
the m=e 28 and 29 peaks remain nearly unchanged,
providing the direct evidence of SiH2 formation
from dissociation of SiH4:
Acknowledgements
We thank Mr. Matt McCormick and Mr. J.
Lemieux for their assistance in this work, and Dr.
Thomas Morton for the useful discussions. Fi-
nancial support from University of California
Energy Institute, National Science Foundation
/CHE-9811400), UC Regents' Faculty Fellowships
and Faculty Development Award, ACS Petroleum
Research Fund /Type G Grant), and the Camille
and Henry Dreyfus Foundation is gratefully ac-
knowledged.
SiH4 ! SiH2 ‡ H2;
which is consistent with previous studies [12±18].
The formation of Si2H6 proceeds via insertion
of silylene into the Si±H bond of silane [12±17]:
ꢁ6†
SiH2 ‡ SiH4 ! Si2H6:
ꢁ7†
The m=e 62 peak increases slightly from 830 to 935
K /Fig. 4), and then declines above 1070 K, indi-
cating both production and pyrolysis of Si2H6. In
addition, from 830 to 1500 K, the appearance and
growth patterns of the m=e 56, 58, and 60 peaks as
a function of temperatures are very similar to
those observed in the pyrolysis of disilane [Figs. 2
and 3 and channels /2)±/5)]. Channel /7), as well as
channels /2)±/5), in silane pyrolysis, is con®rmed
by the production of SiH2 and Si2H6, and by the
pyrolysis product patterns and temperature de-
pendence that are similar to those for Si2H6.
References
[1] J.M. Jasinski, S.M. Gates, Acc. Chem. Res. 24 /1991) 9.
[2] M.S. Gordon, T.N. Truong, E.K. Bonderson, J. Am.
Chem. Soc. 108 /1986) 1421.
[3] P.M. Agrawal, D.L. Thompson, L.M. Ra€, J. Chem. Phys.
92 /1990) 1069.
[4] H.M. Mo€at, K.F. Jensen, R.W. Carr, J. Phys. Chem. 96
/1992) 7683.

488
S.D. Chambreau, J. Zhang / Chemical Physics Letters 343 +2001) 482±488
[5] H.M. Mo€at, K.F. Jensen, R.W. Carr, J. Phys. Chem. 96
/1992) 7695.
[19] J. Boyle, L. Pfe€erle, J. Lobue, S. Colson, Combust. Sci.
Technol. 70 /1990) 187.
[6] J. Dzarnoski, S.F. Rickborn, H.E. O'Neal, M.A. Ring,
Organometallics 1 /1982) 1217.
[20] D.W. Kohn, H. Clauberg, P. Chen, Rev. Sci. Instrum. 63
/1992) 4003.
[7] G. Olbrich, P. Potzinger, B. Reimann, R. Walsh, Organo-
metallics 3 /1984) 1267.
[21] J.A. Blush, H. Clauberg, D.W. Kohn, D.W. Minsek, X.
Zhang, P. Chen, Acc. Chem. Res. 25 /1992) 385.
[22] S.G. Lias, J.F. Bartness, J.F. Liebman, J.L. Holmes, R.D.
Levin, W.G. Mallard, Ion energetics data, in: NIST
Chemistry WebBook, NIST Standard Reference Database,
no. 69, National Institute for Science and Technology,
Gaithersburg, MD, 2000.
[8] J.G. Martin, H.E. O'Neal, M.A. Ring, Int. J. Chem. Kinet.
22 /1990) 613.
[9] J.E. Johannes, J.G. Eckerdt, J. Appl. Phys. 76 /1994) 3144.
[10] J. Simon, R. Feurer, A. Reynes, R. Morancho, J. Anal.
Appl. Pyrolysis 24 /1992) 51.
[11] K. Tonokura, T. Murasaki, M. Koshi, Chem. Phys. Lett.
319 /2000) 507.
[23] B. Ruscic, J. Berkowitz, J. Chem. Phys. 95 /1991) 2407.
[24] B. Ruscic, J. Berkowitz, J. Chem. Phys. 95 /1991) 2416.
[25] B. Ruscic, J. Berkowitz, H. Cho, J. Chem. Phys. 86 /1987)
1235.
[12] J.H. Purnell, R. Walsh, Proc. R. Soc. Ser. A 293 /1966)
543.
[13] M.A. Ring, H.E. O'Neal, J. Phys. Chem. 96 /1992) 10848.
[14] R. Becerra, R. Walsh, J. Phys. Chem. 96 /1992) 10856.
[15] P. Ho, M.E. Coltrin, W.G. Breiland, J. Phys. Chem. 98
/1994) 10138.
[26] D.M. Lubman, R.M. Jordan, Rev. Sci. Instrum. 56 /1985)
373.
[27] S.D. Chambreau, J. Zhang, J.C. Traeger, T.H. Morton,
Int. J. Mass Spectrom. 199 /2000) 17.
[16] H.M. Mo€at, K.F. Jensen, R.W. Carr, J. Phys. Chem. 95
/1995) 145.
[28] J. Zhang, K. Xu, G. Amaral, Chem. Phys. Lett. 299 /1999)
285.
[17] A.A. Onischuk, V.P. Strunin, M.A. Ushakova, V.N.
Pan®lov, Int. J. Chem. Kinet. 30 /1998) 99.
[18] M.T. Swihart, S.L. Girshik, J. Phys. Chem. 103 /1999) 64.
[29] H. Chatham, D. Hills, R. Robertson, A. Gallagher, J.
Chem. Phys. 81 /1984) 1770.