Vacuum 122 (2015) 154e160

Contents lists available at ScienceDirect

Vacuum
journal homepage: www.elsevier.com/locate/vacuum

Experimental study of the frequency factor in the PolanyieWigner

equation: The case of C2H6
R. Luna*, C. Milla
Satorre

n, M. Domingo, C. Santonja, M.A.
Centro de Tecnologías Físicas, Universitat Polit
ecnica de Val
encia, 46022 Val
encia, Spain

a r t i c l e i n f o a b s t r a c t

Article history: The C2H6 molecule has been used to determine experimentally, for the first time, the frequency factor
Received 31 July 2015 present in the PolanyieWigner equation and to study how temperature influences this magnitude for a
Received in revised form zeroth order desorption. This parameter is necessary to calculate the desorption rates for environments
14 September 2015
in which this process occurs. The method presented is based on the analysis of a quartz crystal micro-
Accepted 15 September 2015
balance signal. In the literature the frequency factor is not experimentally obtained but is rather assumed
Available online 21 September 2015
to be kB.T/h (at 50 K), as proposed by the activated state theory for first order desorption processes, or it
is estimated by other methods. Additionally, the factor's variation with temperature has not been
Keywords:
Ice films
experimentally explored to date.
Sublimation energy Two different types of zeroth order desorption experiments have been designed for this study. The
Experimental laboratory purpose of the first experiment, carried out at a constant rate of warming, is to obtain the desorption
energy, which is compared with previous values reported in the literature. The second group of
desorption experiments is performed at constant temperatures and is used to calculate and study the
frequency factor. Several temperatures have been specifically selected, enabling us to determine the
influence of the temperature on this parameter. We have calculated a relationship for the frequency
factor and temperature, obtaining an increase of approximately 50% for the frequency factor for an in-
crease of only 6 K. This result must be taken into account when the PolanyieWigner equation is used for
desorption rate calculations.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction factor, Edes is the desorption energy, T is the absolute temperature

and R is the constant for ideal gases. This topic was further devel-
Desorption experiments are widely performed to study the oped in the works of Redhead [3] and Carter [4].
properties of many materials. From the theoretical point of view, Eq. (1) is used in several fields: catalysis, chemical reactions, as-
the concepts of desorption and sublimation imply different pro- tronomy, desorption kinetics from different material surfaces with a
cesses, however, in the literature concerning low temperature ices, specific orientation, etc. The parameters involved, can be applied to
the two are usually assumed to be equivalent (see for example desorption processes to determine the sticking coefficient, the sur-
Sandford et al. [1]). face coverage, the residence time, or the order of the process for a
The process involved in these experiments is governed by the molecule. The latter is related to the kinetics of a molecule in an
PolanyieWigner equation [2]: elementary process, and is usually an integer i ¼ 0, 1, 2,…, but some
authors, have even obtained a non-integer number [9].
  Another relevant application is to determine the desorption
dN E

¼ Ai $N i $exp
des (1) energy of a species, Edes, which is defined as the heat required to
dt R,T
desorb 1 mol of a substance. This parameter is important in the
where N is the surface coverage of molecules, i is the order of the computation of several thermodynamic magnitudes such as the
process, Ai is the frequency factor, (Ai$Ni) is the preexponential reticular energy of ionic compounds, the energy of formation in the
gas phase from the energy of formation in the condensed phase, the
bonding energy of a molecule with the surface of a specific material,
* Corresponding author. or the sublimation rate rdes of molecules from their bulk ice. In the
E-mail address: ralunam@fis.upv.es (R. Luna).

http://dx.doi.org/10.1016/j.vacuum.2015.09.021
0042-207X/© 2015 Elsevier Ltd. All rights reserved.
 R. Luna et al. / Vacuum 122 (2015) 154e160 155

literature, three procedures to obtain Edes using the Pola- exponential factor for zeroth order, has been calculated as
nyieWigner equation are reported: i) a group of works is based on A0 ¼ A1$N, where N is the surface coverage (molecule cm
2). A value
the use of a linear form of the PolanyieWigner equation by taking of N ¼ 1015 molecule cm
2 is generally used, which implies a surface
the natural logarithm of Eq. (1), then Edes is obtained from the slope with perfectly packed spherical particles with a diameter of 3.2 Å.
of ln(rdes) versus 1/T [10]; ii) in a second group of works, the The entries in column 4 demonstrated how the calculated and
desorption energy is obtained by fitting directly the curve obtained estimated values of Ai vary by several orders of magnitude for the
during thermal programmed desorption (TPD) experiments under same desorption order.
UHV conditions, using Edes as a fitting parameter; in this case, it is Despite the dependence of Ai on the temperature, no experiments
necessary to first estimate a value for the frequency factor (Ai) [11] have been conducted to study this behaviour. Therefore it is worth
or to treat it as an additional fitting parameter [12]; iii) the third studying the influence of temperature on this parameter to know
group of experiments involves desorption at 2 or 3 different con- whether the values typically used are subject to a significant uncer-
stant temperatures implying a constant desorption rate (for each tainty. This study intends to analyse several desorption experiments
temperature); then the desorption energy is obtained from each at different constant temperatures rather than to carry out desorp-
experiment, and the mean Edes is assumed to the representative tion experiments in which the temperature varies continuously.
value of the desorption energy [1]. It is remarkable that for all the To perform these studies, desorbing molecules have been
techniques mentioned, a constant value for Ai has been assumed, measured by three different methods in the literature: i) indirectly,
despite the fact that in most of the processes, the temperature is not by detecting the molecules remaining in the solid phase by means
constant and Ai depends on temperature, as declared in the acti- of an infrared spectrometer [1] and measuring the column density,
vated state theory [13]. Therefore, some authors propose to deter- but when phase changes are involved, this could lead to experi-
mine Ai for each species under the specific conditions of each mental errors; ii) by detecting a fraction of the desorbing molecules
experiment; however, that type of study has not been completed. by using a mass spectrometer, in typical temperature programmed
Due to this lack of studies, one additional application of the desorption (TPD) experiments, carried out under ultra high vacuum
PolanyieWigner equation is to determine how temperature in- conditions [12]; or iii) directly measuring the desorbing molecules
fluences the frequency factor (which has not been experimentally by using the signal of a quartz crystal microbalance (QCMB) during
explored so far). Thus, it is interesting to study this variation with the desorption [16], which implies a more accurate detection of the
temperature to know whether the general assumption, used in the desorbing molecules compared to the previous methods.
literature, a negligible variation of Ai with temperature is In a previous work [16] we have developed a method to calcu-
reasonable. late the desorption energy based on the analysis of the signal
From Eq. (1), it is possible to infer that the units of Ai depends on variation of a QCMB under high vacuum conditions. Using this
the order of the process involved; therefore, the order must be result, it is possible to design a procedure to calculate A0 (frequency
taken into account for the meaning of this parameter. For a first factor) in the PolanyieWigner equation for zeroth order desorption
order desorption process, A1 can be related to the lattice vibrational processes. This parameter is suitable for determining the desorp-
frequency of the molecule within its position in the matrix tion rate necessary to study the desorption process of molecules
[14,15,5]. In fact, A1 is the only Ai whose units are in accord with its present in different environments.
name (“frequency factor”). This is the reason why, in the literature, To explain the method to obtain A0 and its variation with tem-
A1 is taken as a reference to calculate the other Ai by means of the perature for the case of ethane, Section 2 contains a description of
expression Ai ¼ A1$N1
i. the experimental setup, Section 3 presents the experimental pro-
In the case of atoms or small molecules following a first order cedure and results which are subsequently discussed in Section 4,
process, this factor is assumed to follow the expression A1 ¼ kB$T/h and finally, the conclusions are presented in Section 5.
[13], where kB is the Boltzman constant, h is the Planck constant
and T is the absolute temperature. Therefore, this parameter is 2. Experimental setup
temperature dependent. Nevertheless in the literature, for all
molecules, the frequency factor (for a first order process) is The experiments are carried out in a system working under high
assumed to be A1 ¼ 1012 s
1, which implies the assumption of a vacuum (HV) conditions. The main components are a QCMB
temperature of 50 K regardless of whether the process occurs at a (1.0 cm2 surface area) and a temperature controller (Fig. 1).
different temperature, even in the most common experiments in The pressure of the experiment (10
7 mbar) is obtained by
which the temperature changes during desorption. For processes of assembling a pair of turbomolecular pumps backed by their cor-
other orders, this parameter differs from the first order process responding rotary pumps and the help of a closed-cycle He cryostat
value by several orders of magnitude because it is related to N acting as a cryopump. The gas under study is previously placed in a
(surface coverage) and for those cases, it is usually estimated by the prechamber. To control the constant flow of molecules entering the
authors. chamber, two parameters are regulated: the pressure in the pre-
The units of Ai are molecule(1
i) cm2(i
1) s
1 [6] where i is the chamber (measured with a capacitive sensor) and the size of the
order of the process. In Table 1, the units and a summary of the aperture of a needle valve.
values frequently used in the literature are reported, where the pre- The edge of the cryostat is in thermal contact with the QCMB (Q-

Table 1
Values for the frequency factor Ai used in the literature for desorption processes and their corresponding units. N is the surface coverage widely assumed as 1015 molecule cm
2
and A1 ¼ kB$T/h (for first order desorption). Generally a temperature of 50 K is assumed and then, A1 ¼1012 s
1.

Desorption order Ai Reference values Range of values reported in the literature Units

0 A0 1027 1025e1032 a, b, c, d molecule cm
2 s
1

1 A1 1012 1011e1015 a, b, d s
1
2 A2 10
3 10
5e10
1 a, d molecule
1 cm2 s
1
References a: Bisschop et al. [5], b: Acharyya et al. [6]
c: Bolina et al. [7], d: Collings et al. [8]
156 R. Luna et al. / Vacuum 122 (2015) 154e160

Fig. 2. The raw QCMB signal during deposition at constant temperature (left) and
desorption at a constant ramp rate of temperature 1 K min
1(right) for C2H6.

Fig. 1. Disposition of the laser beams and location of the QCMB, the sample and one of
the temperature sensors. obtained by repeating the experiment at least 5 times for a specific
temperature leading to results with a standard deviation lower
than 1%. Next, we obtained the quotient of the density of each ice
Sense gold coated quartz, 5 MHz). The temperature of the sample studied in our laboratory and the corresponding one of Roux et al.
(deposited onto the QCMB) is governed by an ITC 503S intelligent [18], for both ices the quotient indicates a difference between them
temperature controller (Oxford Instruments), using two silicon of less than 1%. This quotient has been used to calibrate and modify
diode sensors (Scientific Instruments), with one located just beside the value of S supplied by the manufacturer of the QCMB to the
the quartz crystal and another beside the heating resistor. This current value used in our experiments under our experimental
arrangement allows the temperature to be varied between 13.0 and conditions. Additionally, with this value we have obtained a density
300.0 ± 0.1 K, by means of a resistor. This controller allows the for other molecules, and we have compared our results obtained at
performance of several types of procedures: i) fixing a constant low temperature with those of Schulze and Abe [19] leading to very
temperature in the QCMB, ii) obtaining a temperature ramp with a similar results.
constant heating rate, and iii) obtaining increasing temperatures During the film growth, when a thickness of approximately 1.0
with selected steps. To measure the film thickness during the micron is achieved (as measured from the interference fringes
growth, interferometric patterns are obtained by means of a double obtained by the laser beam), the deposition is stopped. The second
HeeNe laser (632.8 nm) system. part of the experiment corresponds to a desorption process starting
The following chemical has been used in this research: at approximately 7100 s in Fig. 2 (right), with a heating ramp of
C2H6
99.99 (Praxair). 1 K min
1 as is widely used by other authors for similar processes.
From the signal recorded in the interval from 7100 to 8000 s, it is
3. Experimental procedure and results possible to calculate the desorption energy, but it is first necessary
to improve the original signal obtained due to two effects during
The procedure used in this work to calculate the frequency the experiment: the influence of the temperature on the QCMB
factor in the PolanyieWigner equation and to determine the in- frequency signal and the effect due to the presence of contaminants
fluence of temperature on this parameter for C2H6, is performed in (mainly water). The detailed procedure to remove these effects is
two steps and is based on different types of experiments. explained in detail in Luna et al. [16].
Once the signal is corrected to remove these effects, the
3.1. Desorption energy desorption energy is calculated using the PolanyieWigner
equation.
First, a desorption at a constant rate of warming is performed.  
dN E
The raw signal recorded during the deposition and desorption for
¼ Ai $Ni $exp
des (2)
one experiment with C2H6 is shown in Fig. 2, where the QCMB dt R,T
frequency signal versus time is plotted. Two sections are clearly Taking
dN/dt as the desorption rate (rdes) of molecules from
differentiated in this plot. In the first interval (from 4200 to 5400 s), the QCMB, and assuming a zeroth order desorption process for our
the deposition is performed. The starting point of the deposition is experiments, we obtain a more suitable expression:
at approximately 4500 s, when the frequency signal starts to
 
decrease with a constant slope due to a constant rate of mass E
rdes ¼ A0 $exp
des (3)
deposition onto the QCMB over the sample holder. The Sauerbrey R$T
[17] equation DF ¼
S$Dm shows the relationship between the
QCMB frequency variation DF and the variation of the deposited where rdes and A0 have units of molecule cm
2 s
1. Experimentally,
mass Dm through the constant S of our system. Due to the relevance rdes has been obtained from the derivative of the frequency versus
of this parameter to our results, the value of S supplied by the time for the signal presented in Fig. 2 (right). The corresponding
manufacturer of the QCMB was initially checked by comparing the derived plot is shown in Fig. 3, where rdes is plotted against tem-
density of two different ices grown in our laboratory with the perature. From the peak obtained in this plot, we directly assign the
respective ones of Roux et al. [18]. The density in our laboratory was desorption temperature (under our experimental conditions)
 R. Luna et al. / Vacuum 122 (2015) 154e160 157

Table 2
Parameters obtained for the linear fit to calculate desorption energy (plot of Fig. 4).

Parameter Value Standard uncertainty

Slope ¼ (
Edes/R)
2321 7

Intercept 37.66 0.09
Regression coefficient 0.998

Table 3
C2H6 desorption energy comparison to other previous results from the literature.

This work NIST

1
Tdes(K) Edes (kJ mol ) Tinterval (K) Edes (kJ mol
1)

82.0 ± 0.1 19.3 ± 0.9 90 20.5, a

80e90 22.5, b
a: Bondi [20]
b: Regnier [21]

Fig. 3. Desorption rate for C2H6.

straight line due to possible systematic errors leads to a different
value of the intercept A0, but it does not affect the value of the slope
of the signal recorded. Therefore, it is necessary to obtain this
corresponding to the molecule studied (82 K for C2H6) and from the
parameter in a different way which implies the need to perform
profile of the curve we can confirm the assumption of a zeroth
several desorption experiments at different constant temperatures
order desorption [16] typical of a multilayer desorption.
to obtain the frequency factor for each temperature.
Now, taking the natural logarithm of Eq. (3), a linear expression
Therefore, this first step is suitable for obtaining Edes for C2H6,
is obtained:
but it is not appropriate for determining how temperature in-
fluences A0 for this molecule. Thus, a second procedure must be
Edes 1
lnðrdes Þ ¼ lnðA0 Þ
$ (4) implemented to study A0 as a function of temperature, for which
R T the value of Edes calculated in a previous step, is a necessary
Therefore, a plot of ln(rdes) versus 1/T leads to a straight line parameter.
whose slope is
Edes/R. Fig. 4 represents this final step. From the
slope, and assuming as a first approach (later checked in this work)
3.2. Frequency factor A0 versus temperature
that the A0 temperature dependence does not influence the value of
the energy of sublimation, Edes is derived. In Table 2, we present the
To study the influence of temperature on the frequency factor
values obtained for the fit shown in Fig. 4. The calculated desorp-
A0, an additional set of experiments has been implemented. Each
tion energy is Edes ¼ 19.3 kJ mol
1. In Table 3, we compare our re-
experiment is carried out at a selected constant temperature close
sults with other previously obtained values in the literature. As can
to the sublimation temperature. According to Eq. (3), if the tem-
be observed, there is a good agreement between our result and the
perature is constant, a constant desorption rate is obtained, and
few earlier reported results, obtained with different techniques.
then the QCMB frequency signal corresponds to a straight line of
Despite the assumption by some authors that A0 could be ob-
constant positive slope when the frequency is plotted versus time,
tained from the intercept of this linear fit, this value is not suitable a
as can be seen in Fig. 5 (the interval between the two vertical dotted
priori. The main reason is the impossibility of assigning A0 to a
lines), for the desorption at 68 K. The set of temperatures used to
specific temperature due to the continuous temperature change
perform these experiments have been meticulously selected as
during the experiment, and additionally, any vertical offset of the

Fig. 4. Plot of ln(rdes) versus 1/T (solid line) and the linear fit (dotted line) for C2H6. Fig. 5. Desorption experiment at 68 K for C2H6.
158 R. Luna et al. / Vacuum 122 (2015) 154e160

follows. To obtain a reasonable desorption rate, the temperature Table 5

must be close to 82 K (for C2H6 desorption experiments) to be Parameters obtained for the linear fit to obtain A0 ¼ f(T).

measurable during a typical experiment; however, for a tempera- Parameter Value Standard uncertainty
ture too close to 82 K, few experimental points would be obtained Slope 3.6.1028 0.2.1028
during the short time that the experiment at these temperatures Intercept
2.1.1030 0.1.1030
takes place, and thus, the experimental uncertainties would be Regression coefficient 0.994
larger. Additionally, for temperatures too much lower, a negligible
desorption rate would be obtained, and the experiments could even
4. Discussion
last several days. Therefore considering the previous statements,
our selected experimental desorption temperatures have been 68,
Among other applications, the PolanyieWigner equation is
70, 72 and 74 K (Fig. 3 roughly shows how the desorption rate varies
widely used to calculate the desorption energy for a substance from
with temperature).
its solid phase or from a surface where the molecule is adsorbed.
From the analysis of the signal obtained for each temperature
For this case, in the literature, a negligible variation with temper-
and assuming a constant Edes with temperature, A0 is derived for the
ature for the frequency factor Ai has been assumed but this
corresponding experiment, taking the antilogarithm of the
assumption has not been checked experimentally. Ai has been
following expression:
taken as a constant parameter when the PolanyieWigner equation
is used for calculation purposes, despite the fact that the activated
Edes state theory assets the influence of temperature on Ai. In the pre-
lnðA0 Þ ¼ lnðrdes Þ þ (5)
R$T sent work we have experimentally determined how temperature
influences this parameter for the case of C2H6 desorption.
obtaining We have performed several desorption experiments of C2H6 at
constant specifically selected temperatures (68, 70, 72 and 74 K) to
  obtain A0 from these experiments and to study how A0 depends on
E
A0 ¼ rdes $exp des (6) temperature. As a result, we have obtained an expression for the
R$T
frequency factor A0 ¼ 3.6$1028$T
2.1,1030 (molecule cm
2 s
1) for
where rdes has been calculated experimentally as the temperature range of 68e74 K, in which a reasonable desorp-
tion rate occurs (for laboratory experiments purposes), in the case
of C2H6. In Table 4 we show the values of A0 for the zeroth order

DN
DF Dm 1 desorption process involved when a multilayer desorption occurs.
rdes ¼ ¼ $ $ $N (7)
Dt Dt DF Mr A An increase of approximately 50% is obtained when the tempera-
ture increases from 68 to 74 K.

1
where t is the experimental time, Dm
DF is S (S is the Sauerbrey Using the expression obtained, and assuming that it can be
constant mentioned in Section 3.1) and Mr is the molar mass. extrapolated to the range of 60e80 K (with 82 K the sublimation
Finally temperature for C2H6 under our experimental conditions), it is
possible to calculate the desorption rate for temperatures within
this interval (see Table 6) as well as the interval of time that a zeroth

ðF2
F1 Þ
1 1
rdes ¼ $S $ $NA (8) order desorption for a deposition of C2H6 with a column density of,
t2
t1 Mr
for example, 1.0$1018 molecule cm
2 would take. As can be observed
in column four there is a difference of six orders of magnitude
where F1, F2, t1 and t2 are the corresponding values for the fre-
between the desorption rate at 60 K and the desorption rate at 80 K.
quencies and times at the starting and ending points respectively of
Therefore, for ethane, desorption at temperatures lower than 60 K
the straight line obtained from the QCMB signal during the
could be considered negligible for this type of experiments.
desorption at a constant temperature (see Fig. 5 for the case of 68 K
To compare the influence of A0 on the desorption rate calculated
desorption). Table 4 shows the experimental values obtained for
by the PolanyieWigner equation, in Fig. 6 we have plotted the
this analysis. Column 1 represents the temperature at which the
desorption rate obtained by assuming A0 is constant (solid squares)
experiments have been performed. For each temperature we have
or by taking A0 to be a function of temperature (open circles). For
performed four different experiments and the mean and its esti-
the first case we have chosen the value of A0 obtained at 70 K, and
mated uncertainty has been reported. Column 2 shows the
for the second case, we have used the expression for A0 as a func-
desorption rate measured experimentally, and in column 3 the
tion of temperature obtained above. As can be observed there is a
value of A0 calculated for each temperature is presented. Our values
deviation between the two values that increases as the temperature
are within the range of the values reported by other authors (see
increases. This result, obtained for C2H6, reinforces the idea that the
Table 1).
temperature dependence of A0 must be taken into account for
The linear fit of A0 versus temperature (see Table 5) provides the
theoretical calculations or when calculating the time needed to
expression A0 ¼ 3.6$1028$T
2.1$1030 (molecule cm
2 s
1) for C2H6
perform a typical desorption experiments.
under our experimental conditions.

Table 4
Experimental values obtained for the desorption of C2H6 at different constant temperatures. Column 1 represents the experimental
temperature, column 2 is the calculated desorption rate, and column 3 shows the A0 obtained for each temperature using the Pola-
nyieWigner equation.

Temp. ±0.1 K rdes ¼
DN/Dt 1015 molecule cm
2 s
1 A0 1029 molecule cm
2 s
1

68.0 0.54 ± 0.04 3.6 ± 0.3

70.0 1.70 ± 0.15 4.3 ± 0.4
72.0 5.2 ± 0.6 5.2 ± 0.6
74.0 14.0 ± 2.2 5.7 ± 0.9
 R. Luna et al. / Vacuum 122 (2015) 154e160 159

Table 6
Calculated A0 and rdes at different temperatures and the desorption time for a deposition with a column density of 1.0$1018molecule cm
2.

Temperature ±0.1 K A0 1028 molecule cm
2 s
1 rdes ¼
DN/Dt1016 molecule cm
2 s
1 Desorption time 108 s

60.0 6 ± 10 0.0001 10
65.0 24 ± 11 0.0076 0.13
70.0 42 ± 12 0.17 0.0059
75.0 60 ± 13 2.2 0.00045
80.0 78 ± 14 20 0.000050

for the first time the frequency factor for a zeroth order process
using the case of ethane as an example and to study the influence of
temperature on this parameter in the PolanyieWigner equation. To
this end, an analysis of the QCMB signal for a zeroth order
desorption has been used.
The frequency factor Ai, is a parameter of the PolanyieWigner
equation that is necessary for the calculation of the desorption rates
for any species under conditions suitable for desorption.
For the particular case of zeroth order desorption processes, A0
values in the literature differ by more than 5 orders of magnitude
for the same molecule (1027e1032 molecule cm
2 s
1), depending
on the author. However similar values of Edes are obtained by all of
them using the PolanyieWigner equation based on experiments
where the temperature varies continuously. In all those cases, A0
has been taken to be independent of the temperature. Our work
demonstrates that A0 clearly changes with temperature. Never-
theless, we show how variations of A0 due to temperature only
produce changes of approximately 1% in the value of Edes when the
Fig. 6. Desorption rate comparison, assuming A0 as constant (solid squares) and using
range of temperature is close to the sublimation temperature,
A0 as temperature dependant (open circles) for C2H6. which explains why the previous values of Edes obtained by
different authors are similar. On the other hand, it is necessary to
As discussed previously in Section 3.1, to obtain Edes as a first highlight that our results are relevant for the correct calculation of
step prior to studying the variation of A0 with temperature for the desorption rates because the influence of temperature on A0 must
case of C2H6, it is assumed that the variation of A0 with temperature be taken into account.
does not affect the value of Edes. To evaluate how this assumption Our study has been based on two types of experiments on the
affects this value, we have calculated Edes (see Table 7) for the zeroth order desorption typical of multilayer deposits. From the
temperatures at which we have performed the desorption experi- first one, based on desorption at a constant ramp rate of temper-
ments (68, 70, 72 and 74 K) assuming for all of them the same value ature, we have obtained the desorption energy Edes for C2H6. This
of A0 ¼ 4.2$1029 molecule cm
2 s
1 (corresponding to 70 K), and value has been compared with those previously reported in the
using the expression literature and a good agreement was found. From the second set of
experiments, based on desorption experiments at a constant tem-
Edes ¼ R$TðlnðA0 Þ
lnðDN=DtÞÞ
perature, we have studied the variation of A0 with temperature for
this molecule. As a result, we find a linear variation in the range
where DN/Dt is the experimental value obtained for each temper-
from 68 to 74 K, A0 ¼ 3.6$1028 T
2.1$1030, which implies an in-
ature. As observed, the desorption energy decreases as the tem-
crease of A0 of approximately 50% within this temperature range.
perature increases. Only a difference of approximately 1% is
The results obtained are within the range of the previously reported
obtained for an increase of 6 K. This difference is within the error
results obtained by other methods. Therefore, our results must be
bars. Additionally, because desorption experiments at a constant
taken into account to calculate the desorption rate of a typical
temperature are most likely performed within this temperature
experiment in the laboratory or in an environment where a similar
range (for the reasons given above), for this molecule, the
temperature is present as could be the case for astrophysical sce-
assumption of a constant A0 in the PolanyieWigner equation does
narios or cryogenic samples.
not significantly affect the obtained value of the sublimation
The influence of A0 on the calculation of Edes has also been
energy.
explored; however for this case no significant variation was
detected.
5. Conclusions
Acknowledgements
In this paper we present a procedure to calculate experimentally
This work was supported by the Plan Nacional FIS2013-48087-
Table 7
C2-2-P of the Ministerio de Economía y Competitividad.
Desorption rate Edes calculation at different temperatures, assuming a constant value
for A0 of 4.20$1029 molecule cm
2 s
1.
References
Temperature ±0.1 K rdes ¼
DN/Dt1015 molecule cm
2 s
1 Edes kJ mol
1

68.0 0.54 ± 0.04 19 [1] S.A. Sandford, L.J. Allamandola, Astrophy. J. 417 (1993) 815e825.
70.0 1.70 ± 0.15 19.3 [2] M. Polanyi, E. Wigner, Z. Phys. 33 (1925) 429e434.
72.0 6.0 ± 0.6 19 [3] P.A. Redhead, Vacuum 12 (1962) 203e211.
[4] G. Carter, Vacuum 12 (1962) 245e254.
74.0 14.0 ± 2.2 19.1 €
[5] S.E. Bisschop, H.J. Fraser, K.I. Oberg, E.F. van Dishoeck, S. Schlemmer, Astron.
160 R. Luna et al. / Vacuum 122 (2015) 154e160

Astrophys. 449 (2006) 1297e1309. A108.

[6] K. Acharyya, G.W. Fuchs, H.J. Fraser, E.F. van Dishoeck, H. Linnartz, Astron. [13] A.K. Galwey, M.E. Brown, Thermochim. Acta 386 (2002) 91e98.
Astrophys. 466 (2007) 1005e1012. [14] S.A. Sandford, L.J. Allamandola, Icarus 87 (1990) 188e192.
[7] A.S. Bolina, W.A. Brown, Surf. Sci. 598 (2005) 45e56. [15] V.V. Krongauz, M.T.K. Ling, L. Woo, U. Purohit, Thermochim. Acta 457 (2007)
[8] M.P. Collings, J.W. Dever, H.J. Fraser, M.R.S. McCoustra, Astrophys. Space Sci. 35e40.
285 (2003) 633e659. [16] R. Luna, C. Mill

Satorre, Vacuum 83 (2009) 942e948.
an, C. Santonja, M.A.
[9] W.A. Brown, A.S. Bolina, Mon. Not. R. Astron. Soc. 374 (2007) 1006e1014. [17] G. Sauerbrey, Z. Phys. 155 (1959) 206e222.
[10] R. Luna, M.A.
Satorre, C. Santonja, M. Domingo, Astron. Astrophys. 566 (2014) [18] J.A. Roux, B.E. Wood, A.M. Smith, R.R. Plyer, Defense Technical Information
A27. Center, Report N AEDC-TR-78e81, 1979.
[11] A. Chakradhar, Y. Liu, J. Schmidt, E. Kadossov, U. Burghaus, Surf. Sci. 605 [19] W. Schulze, H. Abe, Chem. Phys. 52 (1980) 381e388.
(2011) 1537e1543. [20] A. Bondi, J. Chem, Eng. Data 8 (1963) 371e381.
[12] G.M. Mun
C. Rogero, C. Atienza,
~ oz Caro, A. Jimnez-Escobar, Martín-Gago JA., [21] J. Regnier, J. Chim, Phys. Phys. Chim. Biol. 69 (1972) 942e944.
S. Puertas, J.M. Sobrado, J. Torres-Redondo, Astron. Astrophys. 522 (2010)