THE JOURNAL OF CHEMICAL PHYSICS 130, 114504 共2009兲
Physical and chemical effects on crystalline H2O2 induced
by 20 keV protons
M. J. Loeffler and R. A. Baragiolaa兲
Laboratory for Atomic and Surface Physics, University of Virginia, Thornton Hall B103,
Charlottesville, Virginia 22904-4238, USA
共Received 25 November 2008; accepted 18 January 2009; published online 17 March 2009兲
We present laboratory studies on radiation chemistry, sputtering, and amorphization of crystalline
H2O2 induced by 20 keV protons at 80 K. We used infrared spectroscopy to identify H2O, O3, and
possibly HO3, measure the fluence dependence of the fraction of crystalline and amorphous H2O2
and of the production of H2O and destruction of H2O2. Furthermore, using complementary
techniques, we observe that the sputtering yield depends on fluence due to the buildup of O2
radiation products in the sample. In addition, we find that the effective cross sections for the
destruction of hydrogen peroxide and the production of water are very high compared to radiation
chemical processes in water even though the fluence dependence of amorphization is nearly the
same for the two materials. This result is consistent with a model of fast cooling of a liquid track
produced by each projectile ion rather than with the disorder produced by the formation of radiolytic
products. © 2009 American Institute of Physics. 关DOI: 10.1063/1.3079612兴
I. INTRODUCTION
Solid hydrogen peroxide, like water ice, is a suitable
substance for fundamental studies of radiation chemistry because of the small number and limited complexity of its radiation products. As seen in our previous studies,1 extended
radiolysis of H2O2 leads mostly to water, O2, O3, and possibly HO2 and HO3.2 Our initial interest in the radiolytic behavior of solid H2O2 stems from the discovery of this molecule on Europa,3 one of the Galilean satellites around
Jupiter, whose surface is composed mostly of water and is
heavily irradiated with ions from the inner magnetosphere.
Recently it was claimed that H2O2 has been detected on the
surface of Enceladus,4 a tiny satellite of Saturn, which is
composed mainly of water ice and the incident radiation is
mostly UV, ions, and electrons. Most icy bodies in the outer
solar system, such as Enceladus and Europa, and grains in
interstellar clouds lack atmospheres that can shield impacts
of energetic particles and photons, and for this reason, we
would expect that hydrogen peroxide is ubiquitous.
Initial studies of pure solid H2O2 have shown that amorphous and crystalline phases can be easily distinguished with
infrared spectroscopy.5,6 Additional studies also showed that
the shapes and positions of some of the absorption bands are
greatly altered when hydrogen peroxide is mixed with
water.6,7 Our more recent study on the 3.5 ␮m band of H2O2
has been applied to Jupiter’s satellite, Europa.7 By comparing our laboratory spectra to the spectrum of the H2O2 on
Europa, we were able to conclude that the molecule is dispersed in the water ice rather than concentrated in precipitates.
There has been a large time gap between the seminal
works on solid hydrogen peroxide5,6 and more recent
ones,1,7–10 probably due to the fact that high concentrations
a兲
Electronic mail: [email protected].
0021-9606/2009/130共11兲/114504/7/$25.00
of H2O2 are currently not commercially available. Thus, to
obtain pure H2O2 or even high concentrations of H2O2, the
sample has to be prepared from the more dilute solutions. In
addition to this obstacle, another problem arises from the
need to study hydrogen peroxide in ultrahigh vacuum 共UHV兲
conditions. The hydrogen peroxide efficiently decomposes
when it contacts any surface of the vacuum chamber and of
the gas-handling manifold. To avoid this problem in our
studies of pure solid hydrogen peroxide in UHV, we have
used two different methods to distill hydrogen peroxide from
a dilute aqueous solution. Both techniques rely on the difference in vapor pressures of these two molecules: We either
evaporate the water after the sample is deposited,9 or we
pump the water out while it is still in the manifold.1,11 The
latter method only works if the manifold is made of a nonreactive material such as glass, while the former can work
with any type of manifold.
The ability to stabilize pure H2O2 in an UHV environment presents an opportunity for a controlled study on radiation effects in pure H2O2. For instance, it allows us to address the question of the stability of the crystalline structure
of solid H2O2 in the presence of irradiation. Ion irradiation is
of particular interest because its high linear energy transfer,
as compared to the case of photolysis,12–14 can have complex
effects on the amount and type of radicals produced in the
film. While previous ion irradiation studies have looked at
destruction rates of hydrogen peroxide,1,8 we are interested
in learning if there is also a connection between ion-induced
amorphization of crystalline H2O2 and the radiolytic formation of new molecules. In the study reported below on the
irradiation of crystalline H2O2 at 80 K with 20 keV H+, we
show how infrared spectroscopy is used to measure amorphization, the production of water, and the destruction of
H2O2 as a function of irradiation fluence 共ions cm−2兲. In previous experiments on the amorphization of crystalline water
130, 114504-1
© 2009 American Institute of Physics
Downloaded 19 Mar 2009 to 128.143.22.132. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/jcp/copyright.jsp
114504-2
J. Chem. Phys. 130, 114504 共2009兲
M. J. Loeffler and R. A. Baragiola
ice,15,16 infrared spectroscopy has proven to be useful for
monitoring the crystalline to amorphous phase transition.
II. EXPERIMENTAL SETUP
All experiments were performed in a stainless steel
vacuum chamber on a radiation-shielded cryostat.1 The base
pressure of the chamber was 4 ⫻ 10−9 Torr and one to two
orders of magnitude lower inside the shield. Hydrogen
peroxide–water mixtures were grown by vapor deposition on
an optically flat gold mirror electrode of a 6 MHz quartzcrystal microbalance 共QCM兲. The areal mass Q 共mass/unit
area兲 of the films was determined by the change in the resonance frequency of the crystal, which was measured with an
Inficon IC/5 controller to a resolution of 0.1 Hz.17 The measured Q can be converted to film column density ␩
共molecules/ cm2兲 if the film composition is known, and converted to thickness if the mass density is known.17
Hydrogen peroxide samples were prepared by depositing
a commercial grade hydrogen peroxide–water mixture
共30 wt % purchased from Alpha Aesar兲 from a stainless steel
manifold onto the microbalance at near normal incidence at
164 K. The relatively high deposition temperature ensured
that impurities, formed when scattered H2O2 molecules react
with the chamber walls, would not stick on the substrate, and
this was verified by infrared spectroscopy and mass spectrometry. After deposition, we used our distillation method9
to obtain crystalline H2O2, where we simply kept the temperature constant and waited for water to evaporate from the
sample. To obtain the desired column density, vapor deposition and distillation were performed twice and the sample
was left to stabilize for ⬃12 h; no decomposition was observed during this time. We chose a column density of H2O2
共7 ⫻ 1017 H2O2 cm−2兲 smaller than the maximum ion penetration depth 共1.1⫻ 1018 H2O2 / cm2 for 20 keV H+ used in
these experiments兲.18
After growth, the films were cooled to 80 K and irradiated with 20 keV protons at near normal incidence, 9°. The
protons were produced by a mass analyzed ion accelerator
and scanned uniformly over the film. A thin wire collector
placed in the ion beam path monitored the proton current and
fluence. The flux on the sample was 2 ⫻ 1012 ions cm−2 s−1.
During irradiation and warming, we monitored the mass
loss of the sample with the QCM, changes in crystallinity
and composition with infrared spectroscopy, and the molecules ejected 共sputtered兲 from the surface using mass spectrometry. The mass spectrometer 共MS兲 was Dycor M200 residual gas analyzer, aimed at near normal incidence to the
sample.
The specular reflectance of the films on the gold substrate was measured at an incidence angle of 35° using a
Thermo-Nicolet Nexus 670 Fourier transform infrared spectrometer operating at 2 cm−1 resolution. The spectra of the
sample were divided by the reflectance of the gold mirror
substrate taken before film deposition. The ratios R共␭兲 were
then converted to optical depth units, −ln R共␭兲. Absorption
band areas were derived after subtraction of baselines that
matched the continuum.
FIG. 1. 共Color兲 Midinfrared spectra of crystalline H2O2 at different fluences
of 20 keV protons: Top 共overtone/combination region兲, middle 共stretching
vibrations兲, and bottom 共bending vibrations兲. The fluctuations at ⬃1.7 ␮m
are due to instrumental noise. The curves, displaced vertically for clarity,
correspond to 共from top to bottom兲 fluences of 0, 1.1, 3.3, 7.5, 20, 77, and
380 in units of 1013 ions/ cm2. H2O2 is abbreviated as HP and H2O is
abbreviated as W.
III. RESULTS
A. Infrared reflectance spectra during ion irradiation
The dependence of the infrared spectra on irradiation
fluence is shown in Fig. 1. As the ion fluence increases, the
sharp features, indicative of crystalline H2O2 transitions,
evolve to broader featureless bands that indicate the amorphous phase.6 This amorphization is accompanied by the appearance of new absorption features, indicative of the synthesis of new molecules. Of the expected products, H2O, H2,
OH, HO2, O2, HO3, and O3, we only can detect H2O, O3, and
possibly HO3 共Fig. 2兲. Furthermore, we do not directly detect
O2 from its forbidden infrared band at 1550 cm−1 above the
noise level but infer its existence in significant amounts from
the presence of the O3 absorption band at 1037 cm−1. The
lack of detection of O2 in the infrared could be related to the
temperature dependence of the absorption coefficient for forbidden transitions, as has been noticed before.19,20
B. Sputtering during ion irradiation
During irradiation we also monitored sputtering from the
sample with the QCM and MS. The sputtering yield, given as
mass loss/ O2 mass, and MS reading of mass 32, the most
abundant product seen during sputtering, are shown in Fig. 3.
As the ion fluence is increased, the sputtering yield and MS
reading at mass 32 are seen to increase dramatically by a
factor of 30–50 from the initial value for the virgin sample,
peaking at ⬃3 ⫻ 1014 ions/ cm2, where it subsequently de-
Downloaded 19 Mar 2009 to 128.143.22.132. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/jcp/copyright.jsp
114504-3
Proton irradiation of crystalline H2O2
FIG. 2. Details of bending vibration region of the infrared spectrum of
hydrogen peroxide after irradiation with 20 keV protons to a fluence of
3.8⫻ 1015 ions/ cm2. Note the bump at ⬃1250 cm−1, which is attributed to
HO3.
creases. The initial sputtering yield has a high statistical uncertainty but we can say that it is a few times larger than the
sputtering yield of water ice 共2H2O / ion兲.21 The correlation
of the MS data with the mass loss measured by the QCM
indicates that the sputtering yield is mainly determined by
the synthesis and release of oxygen into vacuum. We attribute the mass 32 signal to sputtered O2 and not O2 from
cracking of ejected H2O2 共Refs. 1 and 19兲 and O3 共Ref. 22兲
in the MS because the sputtering yield initially increases as
H2O2 decomposes. In contrast, the ozone absorption at
1037 cm−1 does not appear until the sputtering yield has
nearly peaked 共at a fluence of 2 ⫻ 1014 ions/ cm2兲 and increases at higher fluences where the sputtering yield decreases. We note also that the H2O2 and O3 signals are below
the detection limit of the MS. We will return to the question
of the fluence dependence of the sputtering yield and the
comparison with other experiments later in Sec. IV.
C. Thermal desorption after ion irradiation
After irradiating to a fluence of 3.8⫻ 1015 ions/ cm2, at
which the sputtering yield has decreased to ⬃6% of its peak
FIG. 3. 共Color兲 Sputtering yield 共쎲兲 and MS O2 reading 共䊊兲 as a function
of fluence for 20 keV protons on crystalline H2O2 at 80 K.
J. Chem. Phys. 130, 114504 共2009兲
FIG. 4. 共Color兲 Sublimation of a crystalline hydrogen peroxide film irradiated with 20 keV protons to a fluence of 3.8⫻ 1015 ions/ cm2. QCM: Mass
loss rate during warming at 1 K/min. The main structures 共labeled 1, 2, and
3兲 of this curve are assigned to oxygen, water, and hydrogen peroxide with
a small amount of water. MS: Mass 32 signal from the MS. The large rise
above 180 K is O2 from H2O2 desorbing from unirradiated regions and
decomposing off the chamber walls. The shift between the peaks in both
curves is due to a delay in the response of the MS.
value, we performed thermal desorption spectroscopy. This
consisted of warming the sample at 1 K/min, while continuously monitoring its mass with the QCM and the ejected
molecules with the MS. The results, shown in Fig. 4, are
consistent with those from our previous studies on irradiation
of amorphous H2O2 at 20 K.1,19 The mass loss between 80
and 145 K is below the sensitivity of our instruments; the
trapped oxygen remains in the water ice until it quickly desorbs at ⬃155 K. The O2 desorption is correlated with the
crystallization of the water ice and has been observed earlier
in irradiated H2O2 samples1,19 and unirradiated water-oxygen
mixtures.23,24 As with our previous experiments on irradiation of H2O2, the amount of oxygen retained during warming, even at the high fluences of this measurement, is higher
than the amount reported for water-oxygen mixtures just below 130 K.24,25 This enhancement in oxygen retention is attributed to trapping at radiation defect sites in the irradiated
water ice, which are not present in unirradiated water-oxygen
mixtures;26 we note that similar enhanced retention has also
been seen for N2 and H2 products in irradiated mixtures of
ammonia and water.27 At temperatures higher than that of the
oxygen desorption, two more changes in sublimation rate are
observed with the QCM 共Fig. 4兲. These are due to desorption
of H2O bound to other H2O and H2O2 molecules and of pure
H2O2. We note that the peak for H2O bound to H2O2 is not
resolved in these experiments because of the relatively high
heating rate of 1 K/min, five times faster that in our previous
studies.1 Integrating the area under each desorption peak, and
assuming that all of the mass 32 signal belongs to O2, we
estimate that the molecular composition of our sample after
irradiation is 74% H2O 共6.1⫻ 1017 molecules/ cm2兲, 22 %
O2 共1.86⫻ 1017 molecules/ cm2兲, and ⬃4% H2O2 共5.3
⫻ 1016 molecules/ cm2兲. The concentration of trapped oxygen at high fluences is slightly lower than the total oxygen
共O2 + O3兲 concentration measured in our previous experiments at 20 K.1
Downloaded 19 Mar 2009 to 128.143.22.132. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/jcp/copyright.jsp
114504-4
J. Chem. Phys. 130, 114504 共2009兲
M. J. Loeffler and R. A. Baragiola
IV. DISCUSSION
There have been studies on radiolysis of hydrogen peroxide in various concentrations of liquid water by ␥-rays,
x-ray photons, and electrons,12–14,28 in frozen H2O2 : H2O
mixtures29 by UV photolysis and ion irradiation,8 and in pure
solid H2O2 by ion irradiation.1 The initial step of radiolytic
decomposition is believed to be
X + H2O2 → OH + OH,
共1兲
where X is the projectile. One photochemical study 共Ref. 14兲
suggested that the decomposition of hydrogen peroxide
could begin with
X + H2O2 → H2O + O,
共2兲
O + H2O2 → OH + HO2 .
共3兲
The fate of the OH radicals produced by either mechanism
depends on the surrounding environment. In the liquid and
solid phases, where the density of H2O2 is high, the OH
formed can either react with the surrounding molecules or
recombine, a result of the cage effect. This close proximity
allows this chain of reactions,30
OH + H2O2 → H2O + HO2 ,
共4兲
HO2 + H2O2 → H2O + O2 + OH.
共5兲
Thus as long as the OH is near other H2O2 molecules, then
the chain between the product OH 关Eq. 共5兲兴 and reactant OH
关Eq. 共4兲兴 will propagate. The termination step of the chain
reaction has been debated in literature,13,14,31 but it appears
that it could be one of the following two reactions:
HO2 + HO2 → H2O2 + O2 ,
共6兲
2HO2 + H2O2 → 2H2O2 + O2 .
共7兲
Later reactions with hydrogen peroxide or other molecules
formed in the ice can also contribute to the chain of events,
X + H2O → OH + H,
共8兲
H + H2O2 → H2O + OH,
共9兲
H + O2 → HO2 .
共10兲
In addition to being critical in the chain reaction, the HO2
from Eqs. 共3兲 and 共10兲 may be important for the production
of O3. We note that a recent isotopic study on H2 16O and 18O
ice mixtures confirmed that HO2 can be formed via
irradiation.2 However, we do not see this feature here, expected at 1135 cm−1, but in our previous study on irradiation
of amorphous H2O2 at 20 K we did detect a weak feature in
this region, giving support to idea that HO2 is involved in
this chain reaction. It may not be visible in these experiments
because at these higher temperatures a fast reaction to a more
stable molecule may be more likely; we note that previous
experiments also showed that the amount of HO2 formed in
the ice decreases as the irradiation temperature increases,32
although it is still seen at 100 K.
FIG. 5. 共Color兲 Derivative of the infrared spectra as a function of fluence of
20 keV H+ ions. To determine the amount of crystallinity in the sample we
monitored the intensity of the 1380 cm−1 peak. Once the sample is amorphous, the peak intensity goes below the noise level. The decreasing peak
intensity at ⬃1382 cm−1 corresponds to increasing fluences of 0, 0.14, 0.67,
1.1, 3.3, 4.9, 7.5, 11, 20, and 380 in units of 1013 ions/ cm2.
A. Effective rate of water production and
amorphization
To understand amorphization, we consider that the protons deposit the energy along their path at an average rate of
⬃40 eV/ nm,18 which is sufficient to melt the region close to
the ion path.33 This melted region cools down and solidifies
extremely fast, of the order of 10 ps, a condition that leads to
the amorphous state. To estimate the radial extent of the
amorphous region, we use the standard defect overlap
model,34 where the relative concentration of the defects 共assumed to be proportional to the amorphous fraction f a兲 is
given by
f a = 1 − exp共− AF兲,
共11兲
where A is the transverse area of the amorphous region produced by each ion and F is the ion fluence 共ions/ cm2兲. To
obtain f a from the infrared data 共Fig. 1兲 we use the fact that
the H2O2 bands are sharp when the sample is crystalline and
broaden as it becomes amorphous. Thus to evaluate f a versus
irradiation fluence we determined the sharpness of the
stretching bands near 3 ␮m and the bending bands around
7 ␮m by taking the derivative of each spectrum and comparing peak intensities during irradiation 共Fig. 5 shows the
7 ␮m region兲. Using the derivative spectra we define f a = 1
− I p共F兲 / I p共0兲, where I p is the peak intensity of the derivative
and F is the fluence. We fit the data with Eq. 共11兲. The
results, averaged for both 3 and 7 ␮m regions, are shown in
Fig. 6; the data is well fit by the defect overlap model with
an offset that suggests that, before irradiation, the sample has
an initial content of amorphous material of 5%. We obtain
A = 共3.8⫾ 0.7兲 ⫻ 10−14 cm2 which, interestingly, matches the
value reported for amorphization of water ice with 30 keV
protons at 16 K.16 This agreement suggests that the size of
the ion track is similar in both cases and that amorphization
Downloaded 19 Mar 2009 to 128.143.22.132. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/jcp/copyright.jsp
114504-5
J. Chem. Phys. 130, 114504 共2009兲
Proton irradiation of crystalline H2O2
FIG. 6. 共Color兲 Fraction of amorphous hydrogen peroxide vs 20 keV H+
fluence, showing a dependence f a = 0.05+ 0.95共1 − exp共−AF兲兲, with A
= 共3.8⫾ 0.2兲 ⫻ 10−14 cm2 and an initial amorphous content of 5%. Also
shown is the H2O content; it follows y = 0.74关1 − exp共共−␴w + ␴ p兲F兲兴, with
␴w + ␴ p = 4.7⫾ 0.2⫻ 10−15 cm2. Error bars represent uncertainties in the
analysis of infrared absorption.
does not depend strongly on radiation chemistry.
It is interesting to analyze how the amorphization rate
compares with the formation rate of the most important products in the sample: H2O and O2. We analyzed the area of the
water absorption bands at ⬃5143 and 1650 cm−1; the results
are shown in Fig. 6 after being normalized to 0.74, which is
the concentration of water in the sample determined using
the QCM and MS data during thermal desorption 共see
above兲. The procedure could not be applied to O2 since its
absorption was not seen in the infrared spectrum at
1550 cm−1 共see above兲.
A simple model that includes the production of water
with a cross section ␴w and the conversion of water to hydrogen peroxide with a cross section ␴ p leads to the rate
equation,
df w
= ␴w共1 − f w兲 − ␴ p f w ,
dF
共12兲
where f w and 共1 − f w兲 are the fraction of water and hydrogen
peroxide, respectively, and where we neglected the small
oxygen fraction. Since the destruction of hydrogen peroxide
and the creation of water occur not with a single collision but
in a chain reaction, the cross sections should be interpreted
as effective areas perpendicular to the ion beam path where
the molecules are destroyed or created. The solution to
equation12 is given by
fw =
␴w
关1 − exp共− 共␴w + ␴ p兲F兲兴.
␴w + ␴ p
共13兲
Figure 6 shows that the experimental data of the water
fraction f w are well represented by this equation with ␴w
+ ␴ p = 共4.7⫾ 0.2兲 ⫻ 10−15 cm2 and ␴w / 共␴w + ␴ p兲 = 0.74 from
the microbalance data 共see above兲. This yields ␴w
= 共3.7⫾ 0.4兲 ⫻ 10−15 cm2 and ␴ p = 共1.0⫾ 0.4兲 ⫻ 10−15 cm2.
We note that the value calculated here for ␴w is an underestimate because oxygen is neglected in our model but included in our concentration 共0.74兲. Thus, we can calculate a
more accurate value of ␴w from d␩共H2O兲 / dF, the initial
FIG. 7. 共Color online兲 Destruction of H2O2 as a function of fluence overlayed with a smoothed fit to the sputtering yield given in Fig. 3. The H2O2
abundance was calculated using the absorption band at 4690 cm-1. The
inset shows the linear fit at low fluences used to derive the destruction cross
section.
slope of the water production curve which, at low fluences, is
equivalent to ␴w ⫻ ␩0共H2O2兲, where ␩0共H2O2兲 is the initial
column density of hydrogen peroxide. Using this procedure,
we obtain ␴w = 共4.1⫾ 0.1兲 ⫻ 10−15 cm2 and then ␴ p
= 共0.6⫾ 0.3兲 ⫻ 10−15 cm2. One can notice that the cross section for producing water, the most abundant radiolytic product, is about approximately ten times smaller than A, the area
of the amorphous track. This large factor shows that the
structural damage produced by dissociation is relatively unimportant for amorphization, explaining why the amorphization rate of water ice and highly reactive hydrogen peroxide
can be so similar.
B. Destruction rate of H2O2
As noted earlier in Sec. IV, the destruction of H2O2 proceeds as a chain reaction, where some of the intermediate
reactive products can be also destroyed. To quantify this effect, we quantified the abundance of H2O2 versus fluence
共Fig. 7兲 in a similar manner as with the water. We chose to
use the H2O2 absorption band at 4690 cm−1 since our blank
experiments indicated that the band area of this H2O2 absorption feature only changed ⬃10% during crystallization.
The initial destruction rate of H2O2, ␴d, calculated from the
initial slope with a linear fit 共inset in Fig. 7兲, is 共4.5⫾ 0.2兲
⫻ 10−15 cm2. The result that ␴d ⬇ ␴w means that, after all
radicals react, the destruction of hydrogen peroxide leads
mainly to H2O + 21 O2—rather than H2 + O2, suggesting that
reactions 共1兲, 共4兲, and 共5兲 are the most important.
To compare this result to our previous experiments at 20
K,1 we introduce the initial radiation yield G, a commonly
used quantity in radiation chemistry, defined as the number
of products produced or destroyed per 100 eV of energy
absorbed. For the destruction of hydrogen peroxide G共–H2O2兲 = −␴d ⫻ ␩0共H2O2兲 / 共⌬E / 100 eV兲 = 19.7⫾ 4,
where ⌬E = 16 keV is the energy absorbed in the film obtained from SRIM,18 which has ⬃20% uncertainty. Thus,
G共–H2O2兲 is similar to the value of 21⫾ 5 measured with 50
Downloaded 19 Mar 2009 to 128.143.22.132. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/jcp/copyright.jsp
114504-6
J. Chem. Phys. 130, 114504 共2009兲
M. J. Loeffler and R. A. Baragiola
TABLE I. Summary of cross sections and radiation yields derived from experiments.
H 2O 2
H 2O
A
共10−15 cm2兲
␴d
共10−15 cm2兲
38⫾ 7
4.5⫾ 0.2
␴w
共10−15 cm2兲
␴p
共10−15 cm2兲
G共H2O兲共100 eV兲−1
4.1⫾ 0.4
0.6⫾ 0.3
17.9⫾ 3.6
G共–H2O2兲共100 eV兲−1
19.7⫾ 4.0
keV protons on amorphous peroxide at 20 K 共Ref. 1兲
but substantially higher than G共−H2O兲 ⬃ 0.5 for water ice at
73 K.35 Similarly we can calculate the initial radiation
yield
for
the
production
of
water,
G共H2O兲
= ␴w ⫻ ␩0共H2O2兲 / 共⌬E / 100 eV兲 = 17.8⫾ 3.6, slightly higher
than the value of 12⫾ 3 measured with 50 keV protons on
amorphous H2O2 at 20 K.1 A summary of cross sections and
G-values are given in Table I.
C. Sputtering of O2
The peak sputtering yield observed in these experiments
is much higher than that measured in previous keV proton
sputtering experiments on solid H2O at 80 K 共Ref. 36兲 and
on solid O2 at 5 K.37 While H2O is compositionally similar
to H2O2, the amount of O2 produced in irradiation of water
ice is at least an order of magnitude lower for proton irradiation of water ice than in our experiments, which translates
into a much lower sputtering yield. The peak sputtering yield
in our experiments is approximately three times higher than
measured on solid O2.37 This difference can be explained by
the fact that the temperature in our experiments is much
higher than the sublimation temperature of solid O2.
The fluence dependence of the sputtering yield can be
understood more easily if we overlay it on top of the destruction of hydrogen peroxide with fluence 共Fig. 7兲. As the hydrogen peroxide is destroyed, oxygen is being created in the
film and sputtered. The high sputtering of ⬃140 produced by
a narrow ion track shows that sputtering is not limited to the
surface but also extracts O2 from the bulk. This suggests that
the concentration of oxygen maximizes at the fluence where
the H2O2 signal disappears. With further irradiation the
amount of O2 decreases, as no more O2 is being created in
the ice.
V. CONCLUSIONS
In this work, we have studied the amorphization and
radiation chemistry of crystalline H2O2 induced by 20 keV
protons at 80 K. Using the combination of the QCM, MS,
and IR, we identify that H2O, O2, and O3 are the main stable
products formed during irradiation, while we also detect a
weak feature that may be due to HO3. Furthermore, we also
observed a fluence dependence on the sputtering yield; initially the yield is low, but as H2O2 decomposes it increases
and peaks after a fluence of ⬃3 ⫻ 1014 ions/ cm2, and then
subsequently decreases. Our mass spectrometry results show
that this fluence dependence of sputtering is driven by the
formation of O2 in the sample and subsequent depletion of
the H2O2 parent species. By warming the irradiated film, we
found that the trapped oxygen remains in the ice until
⬃155 K, where H2O crystallizes. Finally, we also deter-
mined quantitatively the cross section for water production
and for H2O2 destruction. Both are very large compared to
those for radiolytic reactions in water ice which we explain
to be due to the chain reaction involving radicals that occurs
in H2O2 during irradiation and which is absent in water. In
contrast, the rates of amorphization of both ices are nearly
the same, showing that they are unrelated to chemical processes. Rather, the amorphization rates reflect the similarity
in the energy deposition profiles around the ion track, and
thus in the extent of the melted region around the projectile
track which, after ultrafast cooling, preserve the molecular
disorder.
ACKNOWLEDGMENTS
This material was based on work supported by the National Science Foundation under Grant No. AST0807830 and
NASA Outer Solar System Program under Grant No.
NNX07AL48G.
1
M. J. Loeffler, B. D. Teolis, and R. A. Baragiola, J. Chem. Phys. 124,
104702 共2006兲.
2
P. D. Cooper, M. H. Moore, and R. L. Hudson, J. Phys. Chem. A 110,
7985 共2006兲.
3
R. W. Carlson, M. S. Anderson, R. E. Johnson, W. D. Smythe, A. R.
Hendrix, C. A. Barth, L. A. Soderblom, G. B. Hansen, T. B. McCord, J.
B. Dalton, R. N. Clark, J. H. Shirley, A. C. Ocampo, and D. L. Matson,
Science 283, 2062 共1999兲.
4
S. F. Newman, B. J. Buratti, R. Jaumann, J. M. Bauer, and T. W. Momary, Astrophys. J. 670, L143 共2007兲.
5
R. L. Miller and D. F. Horning, J. Chem. Phys. 34, 265 共1961兲.
6
P. A. Giguere and K. B. Harvey, J. Mol. Spectrosc. 3, 36 共1959兲.
7
M. J. Loeffler and R. A. Baragiola, Geophys. Res. Lett. 32, L17202
共2005兲.
8
R. L. Hudson and M. H. Moore, Astrobiology 6, 483 共2006兲.
9
M. J. Loeffler, B. D. Teolis, and R. A. Baragiola, J. Phys. Chem. B 110,
6911 共2006兲.
10
N. Boudin, W. A. Schutte, and J. M. Greenberg, Astron. Astrophys. 331,
749 共1998兲.
11
M. J. Loeffler, Ph.D. thesis, University of Virginia, 2007.
12
F. S. Dainton and J. Rowbottom, Nature 共London兲 168, 1160 共1953兲; E.
R. Johnson, J. Chem. Phys. 19, 1204 共1951兲.
13
E. J. Hart and M. S. Matheson, Discuss. Faraday Soc. 12, 169 共1952兲; J.
Weiss, ibid. 12, 161 共1952兲.
14
J. P. Hunt and H. Taube, J. Am. Chem. Soc. 74, 5999 共1952兲.
15
R. A. Baragiola, M. J. Loeffler, U. Raut, R. A. Vidal, and C. D. Wilson,
Radiat. Phys. Chem. 72, 187 共2005兲; G. Strazzulla, G. Leto, G. A.
Baratta, and F. Spinella, J. Geophys. Res. 96, 17547 共1991兲; G. Strazzulla, G. A. Baratta, G. Leto, and G. Foti, Europhys. Lett. 18, 517 共1992兲.
16
G. Leto and G. A. Baratta, Astron. Astrophys. 397, 7 共2003兲.
17
N. J. Sack and R. A. Baragiola, Phys. Rev. B 48, 9973 共1993兲.
18
J. F. Ziegler, Stopping and Range of Ions in Matter SRIM2003 共available at
www.srim.org兲.
19
M. J. Loeffler, B. D. Teolis, and R. A. Baragiola, Astrophys. J. 639, L103
共2006兲.
20
B. R. Cairns and G. C. Pimentel, J. Chem. Phys. 43, 3432 共1965兲.
21
M. Shi, R. A. Baragiola, D. E. Grosjean, R. E. Johnson, S. Jurac, and J.
Schou, J. Geophys. Res. 100, 26387 共1995兲.
22
M. Famá, D. A. Bahr, B. D. Teolis, and R. A. Baragiola, Nucl. Instrum.
Methods Phys. Res. B 193, 775 共2002兲.
Downloaded 19 Mar 2009 to 128.143.22.132. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/jcp/copyright.jsp
114504-7
23
J. Chem. Phys. 130, 114504 共2009兲
Proton irradiation of crystalline H2O2
M. P. Collings, M. A. Anderson, R. Chen, J. W. Dever, S. Viti, D. A.
Williams, and M. R. S. McCoustra, Mon. Not. R. Astron. Soc. 354, 1133
共2004兲.
24
R. A. Vidal, D. Bahr, R. A. Baragiola, and M. Peters, Science 276, 1839
共1997兲.
25
R. A. Baragiola and D. A. Bahr, J. Geophys. Res. 103, 25865 共1998兲.
26
B. D. Teolis, R. A. Vidal, J. Shi, and R. A. Baragiola, Phys. Rev. B 72,
245422 共2005兲.
27
M. J. Loeffler, U. Raut, and R. A. Baragiola, Astrophys. J. 649, L133
共2006兲.
28
M. Ebert and J. W. Boag, Discuss. Faraday Soc. 12, 189 共1952兲.
29
V. S. Gurman, V. A. Batyuk, and G. B. Sergeev, Kinet. Katal. 8, 527
共1967兲.
F. Haber and R. Willstatter, Ber. Deut. Chem. Ges. B 64, 2844 共1931兲.
W. C. Schumb, Hydrogen Peroxide 共Reinhold, New York, 1955兲.
32
P. D. Cooper, M. H. Moore, and R. L. Hudson, Icarus 194, 379 共2008兲.
33
G. Szenes, Radiat. Eff. Defects Solids 162, 557 共2007兲.
34
J. F. Gibbons, Proc. IEEE 60, 1062 共1972兲.
35
J. W. T. Spinks and R. J. Woods, An Introduction to Radiation Chemistry
共Wiley, New York, 1990兲.
36
R. A. Baragiola, R. A. Vidal, W. Svendsen, J. Schou, M. Shi, D. A. Bahr,
and C. L. Atteberrry, Nucl. Instrum. Methods Phys. Res. B 209, 294
共2003兲.
37
M. Famá, B. D. Teolis, D. Bahr, and R. A. Baragiola, Phys. Rev. B 75,
100101共R兲 共2007兲.
30
31
Downloaded 19 Mar 2009 to 128.143.22.132. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/jcp/copyright.jsp