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

Подождите немного. Документ загружается.

120 Charged Particle and Photon Interactions with Matter

decreases from 240 to 280 eV as the photon energy increases. Two maxima can be seen at about 290

and 305eV, accompanied by a minimum at about 295eV. From 310 to 400eV, the W-value remains

essentially constant.

An electron energy loss spectrum and a photoabsorption spectrum of methane show the four

peaks at 287, 288, 289.4, and 289.8eV (Brown et al., 1978; Wight and Brion, 1974). A strong reso-

nance is located at 288eV, and the three other peaks are weak in intensity. The transitions of an

electron from the inner C K-shell to Rydberg orbitals give rise to these peaks. The K-shell ioniza-

tion energy is 290.8eV (Pireaux et al., 1976). The oscillatory structure shown by the circles in

Figure6.11

is presumed to connect with the structure caused by transitions of the K-shell electron.

model, which is the same as that for ethylene, seems applicable for interpreting the structure

in the energy dependence of the W-value (Saito and Suzuki, 1986; Suzuki and Saito, 1987). Since

methane has no transition into the π* orbital, the oscillatory structure is simpler than that for

ethylene. Then, the energy region of methane classied by the photoabsorption process is 5, which

is smaller than that for ethylene. The W-value calculated is illustrated with a solid curve in Figure 6.11.

The calculation has reproduced the measured data essentially well, although it overestimates the

effect on PCI just above the C K-edge.

Propane has K-shell ionization energy of 290.6eV (Pireaux et al., 1976). Transitions of the 1s

electrons to unoccupied orbitals are supposed to take place slightly below the ionization threshold.

These types of excitation and ionization probably cause molecules to change into different types

of states of the molecular ion. The W-value is expected to be affected by these ion states near the

K-edge. The relative W-values derived from the measurements are shown with open triangles for

photon energies from 260 to 360eV in Figure 6.12 (Suzuki and Saito, 1985b). The solid curve indicates

a calculated prole based on the model mentioned below. The measured W-value slightly decreases

with increasing energy from 260 to 280 eV. The W-value shows a sharp rise at 286eV, goes to

a maximum at 291 eV, and then decreases to a minimum at 294eV. A linear dependence of the

W-value data on the photon energy is seen from 294 to 304 eV. The W-value is essentially constant

above

304

eV.

In order to clarify the origin of the variation in the W-value shown in Figure 6.12, Suzuki and

Saito (1985b) observed the photoabsorption spectrum of propane around the C K-edge. Although

the inner-shell excited states are not separated from one another owing to the low resolving power of

the monochromator used, the following items are possibly discernible. The absorption efciency is

260

1.00

1.04

1.08

Relative W-value

280 300

Photon energy (eV)

320 340 360

Figure 6.12 Photon W-value for C

as a function of soft x-ray energy. Open triangles indicate mea-

sured data. The solid curve denotes the calculated prole based on the model explained in Section 6.4.5.

(Reproduced

from Suzuki, I.H. and Saito, N., Bull. Chem. Soc. Jpn., 58, 3210, 1985b.)

New Directions in W-Value Studies 121

low below 285eV, becomes high at about 287eV, and reaches a maximum at around 291eV. Above

291eV, this efciency gradually decreases with increasing photon energy. Judging from the inner-

shell excitation spectra of methane and ethane (Eberhardt et al., 1976), the high efciency at 287eV

can probably be ascribed to a transition of C-atomic orbitals, 1s → 3s. In the cases of methane and

ethane, the transition (1s → 3s) is low in intensity. This type of transition is supposed to occur with

a higher probability in propane because of the poorer symmetry than those for methane and ethane.

This transition is followed by transitions to higher excited states (e.g., 1s → 3p) and to ionized states.

Bearing in mind the photoabsorption spectrum, a model similar to that for methane is proposed for

the interpretation of the energy dependence of the W-value (Suzuki and Saito, 1985b). The calculated

curve

has well reproduced the measured data and the model is essentially correct.

6.5 osCillatory variation in energy dependenCe oF photon

W-values

are

toms

and a

tomiC shell

eFF

eCts

6.5.1 introduction

Based on the progress in synchrotron radiation researches, many photoabsorption spectra have

been measured in a high resolution of photon energy, and electron emission spectra have also been

observed in a high resolution of electron energy (Svensson et al., 1988; Chen et al., 1989; Okada

et al., 2005; Kato et al., 2007). Further, several ions with various charges, produced by the photo-

ionization of atoms and molecules, have been observed by means of mass spectrometry (Hayaishi

etal., 1984; Saito and Suzuki, 1992, 1994, 1997; Tamenori et al., 2004; Suzuki et al., 2006). In some

of these studies, average charge states have been obtained for ions produced from rare gas atoms

irradiated with soft x-rays (Saito and Suzuki, 1992, 1994; Tamenori et al., 2004). Some measure-

ments were attempted for the determination of the absolute intensity of soft x-rays using a multi-

electrode ion chamber and other techniques as per the requirements in photon science and photon

engineering (Rabus et al., 1997; Saito and Suzuki, 1998, 1999). Photoionization cross sections of

rare gas atoms were obtained on an absolute scale in the soft x-ray region using a sophisticated ion

chamber technique (Suzuki and Saito, 2002, 2005). These developments have enabled us to deter-

mine absolute photon W-values of some gases in the sub-keV energy regions, even though it is very

difcult

for the proportional counter technique to provide accurate W-values on the absolute scale.

this section, photon W-values for rare gas atoms are presented, which were obtained using

monochromatic synchrotron radiation combined with the multielectrode ion chamber (Saito and

Suzuki, 2001a,b; Suzuki and Saito, 2001). In deriving these quantities from observed photoion cur-

rents, average charge states, which are called γ-values here, have been utilized. Branching ratios of

produced ion charges in individual photon energies are used in the calculation of the W-value on the

basis of a precise model, which takes atomic shell effects into consideration. First, a measurement

method and an analysis technique are given, and the result for Kr is discussed in detail in compari-

son with a calculated prole (Saito and Suzuki, 2001b). Then, results for Ar and Xe are presented

(Saito

and Suzuki, 2001a; Suzuki and Saito, 2001).

6.5.2 MeaSureMentS

Synchrotron radiation at the AIST (National Institute of Advanced Industrial Science and

Technology) was dispersed by a Grasshopper grating monochromator (Saito and Suzuki, 2001a,b;

Suzuki and Saito, 2001). The monochromatic photons entered the multielectrode ion chamber,

whose shape is cylindrical, 65mm in diameter and 1300mm in length. In order to reduce higher-

order radiation and scattered lights, the electron energy of the storage ring was often lowered from

750MeV to 250–600MeV, and some thin lms were inserted into the soft x-ray beam. Figure 6.13 shows

the schematic diagram of the measurement system for W-values (Saito and Suzubi, 2001b). The

window of the ion chamber consists of a thin VYNS foil and a circular aperture of 2mm diameter.

122 Charged Particle and Photon Interactions with Matter

The chamber contains a set of six electrodes for detection of photoion currents and for suppression

of end distortion in electric elds. The sample gases of Kr, Ar, and Xe (research grade purity) were

supplied at about 10

−2

–10

Pa, the densities of which were monitored with a capacitance manometer

and controlled with an automatic valve system. A positive potential was applied to the outer cylin-

drical electrode. The photoion current at each inner electrode was transferred to a pico-ammeter and

then to a personal computer. The incident intensity of the soft x-rays was continuously monitored

with

the Au-plated Ni-mesh, illustrated in Figure 6.13.

6.5.3 analySiS

Let us consider that a total of N electrons are produced upon absorption of an incident soft x-ray of

energy E

in a sufciently high density gas (Saito and Suzubi, 2001b). Then, the photon W-value,

, is expressed by

= . (6.9)

The number of electrons, n, produced in the ion chamber with a certain gas density, p, is the sum

of contributions from two factors. The rst factor is the initial ionization through atomic multiple

photoionization effects, and the second is the secondary ionization effects originating from colli-

sions

of emitted electrons with ambient gases in the chamber. Thus,

n p p( ) ( ),= +γ δ (6.10)

where

denotes the number of electrons ejected from the atom having absorbed a soft x-ray quantum

δ is the number of electrons secondarily produced

The

value of γ was determined previously for several rare gas atoms in the soft x-ray region using

a time-of-ight mass spectrometer technique (Saito and Suzuki, 1992; Suzuki and Saito, 1992).

Monochromator

Differential

pumping

system

Filter

Photon

monitor

Aperture

window

Capacitance

manometer

Soft x-ray beam

Ion chamber

Pico-ammeter

Personal

computer

Turbo-molecular

pump

Gas supply

Pressure

controller

Photon

monitor

Figure 6.13 Schematic of an experimental system for measurements of the photon W-value using a multi-

electrode ion chamber. The personal computer controls the monochromator, the voltage supply (+V), and the

pressure

controller. (Reproduced from Saito, N. and Suzuki, I.H., Radiat. Res., 156, 317, 2001b.)

New Directions in W-Value Studies 123

Asthe gas density decreases, the parameter value δ approaches zero and the n value becomes equal

to the γ value. On the other hand, the δ value approaches a constant value and the n value reaches a

plateau value of N, as the gas density increases. An example of results measured at a photon energy

of 400 eV is shown in Figure 6.14, where the number of electrons produced is represented as a func-

tion of the Kr density (Saito and Suzubi, 2001b). From the curve in Figure 6.14, constant values at

low- and high-density limits are easily estimated with a high precision. When the purity of the soft

x-ray beam was not very high, the plateau value did not show up clearly. This situation has been

overcome

by the improvement of the photon purity, as indicated above.

described previously on the absolute measurement of soft x-rays (Saito and Suzuki, 1998,

1999),

the photoion current, i, in the ion chamber under a certain gas density is given as

i enI l p L p= − − −

{ }

exp( ) exp( ) .σ σ1 (6.11)

Here, e, I, σ, L, and l denote the elementary charge, the photon absolute intensity, the photoabsorp-

tion cross section, the length of the electrode, and the length of the insensitive region at the front

end,

respectively. Then, the photoion current at a sufciently low gas density is

i e I l p L p= − − −

{ }

γ σ σexp( ) exp( ) .1 (6.12)

Similarly,

the photoion current at a sufciently high density is

i eNI l p L p= − − −

{ }

exp( ) exp( ) .σ σ1 (6.13)

The photoabsorption cross section has been obtained with ion currents from the two successive

electrodes,

and i

, under appropriate gas densities as

σ =

























ln . (6.14)

By using Equations 6.12 and 6.14, we have obtained the absolute intensity of the incident soft x-ray.

Then, the value of N, that is, the total number of electrons produced, has been derived from the

0.1 1 10

100 1000

Number of electrons produced

Kr density (Pa)

Figure 6.14 Number of electrons produced in the ion chamber as a function of Kr gas density. Data are

plotted at a photon energy of 400.3eV. See Section 6.5.3 for N and γ. (Reproduced from Saito, N. and Suzuki,

I.H.,

Radiat. Res., 156, 317, 2001b.)

124 Charged Particle and Photon Interactions with Matter

obtained photon intensity using Equation 6.13. Finally, the photon W-value has been calculated from

the

derived N value, according to Equation 6.9.

Figure

6.15 shows the value of γ for Kr in the region of 60–1100eV (Saito and Suzuki, 2001b).

These values have been derived from branching ratios into several charge states in multiple pho-

toionization (Suzuki and Saito, 1992). It is found that the γ-value shows a steep jump at the 3d

electron ionization threshold (93.8eV). This nding can be interpreted in terms of the formation of

doubly charged Kr ions through the 3d electron ionization. Even below the 3d threshold, the γ-value

is slightly higher than unity, showing that two valence electrons are appreciably ejected through

absorption of one photon. The value of γ increases gradually above the 3d threshold, and then

exhibits small upward steps at the 3p (214.4eV) and 3s (292.8eV) electron ionization thresholds.

Above the 3s threshold, the γ-value acquires an additional feature. This feature probably comes

from a shake-off process induced by the inner-shell ionization (Saito and Suzuki, 1994, 1997;

Suzuki et al.,

2006).

6.5.4 reSult of kr

The photon W-value of Kr is shown on an absolute scale as a function of photon energy from 85 to

1000eV in Figure 6.16 (Saito and Suzuki, 2001b). The present data are denoted with solid circles.

The bars with hatching indicate the ionization thresholds of 3d, 3p, and 3s electrons. The open

square data points give the W-value for electrons, W

, which were reported earlier and are shown

in Figure 6.1 (Combecher, 1980). The solid curve represents the result calculated from the model

described below. The measured W

exhibit considerably lower values than those for electrons

below 91eV. The sharp peak structure appears just below the 3d threshold, which is shown in

Figure 6.17.

The value of W

steeply increases at the 3d ionization threshold and decreases slightly between

120 and 210eV. This W

shows a small oscillation near the 3p and 3s ionization thresholds, and

then essentially reaches a constant value above 400eV. The W

for electrons is close to the present

measured value of W

in energies between 200 and 300eV. The variation in W

near the ionization

thresholds is supposed to originate from inner-shell ionization effects, which is similar to those for

hydrocarbon molecules near the C 1s ionization thresholds, as described in Section 6.4.4 (Suzuki

and Saito, 1985a,b, 1987; Saito and Suzuki, 1986). Then, the mechanism underlying these phenom-

ena

is assumed to have a close relation with the model proposed in regard to these molecules.

100 1000

γ-value

Photon energy (eV)

Figure 6.15 γ-Value (average charge state) of Kr as a function of photon energy. Bars with hatching denote

the ionization thresholds of 3d, 3p, and 3s electrons. (Reproduced from Saito, N. and Suzuki, I.H., Radiat.

Res.,

156, 317, 2001b.)

New Directions in W-Value Studies 125

6.5.5 preciSe Model of the atoMic Shell effect on variation in the W-value of kr

The model for explaining the oscillatory variation in the measured W

value is described as follows

(Saito and Suzuki, 2001b): The number of secondary electrons produced by the electrons emitted

from

a Kr atom is calculated using the experimental data of Combecher (1980).

(i) Below the 3d threshold: Single, double, and triple valence photoionizations take place.

In the case of single ionization, a valence electron of 4p or 4s with an energy of E

, respectively, is ejected (Svensson et al., 1988; Schmidt, 1997). The numbers of elec-

trons produced from collisions of the 4p and 4s photoelectrons with ambient atoms can

be calculated to be E

) and E

), respectively. The branching ratio of the

4p and 4s ionizations is simply assumed to be the multiplicity of the two orbitals, that is,

3:1. Then the total number of electrons produced by single valence ionization, N

, is given

90 92 94 96

3/2

5/2

W-value

Photon energy (eV)

Figure 6.17 Photon W-value for Kr in the region of 3d thresholds as a function of photon energy. Solid

circles show measured data. The bars with hatching denote ionization thresholds for 3d

3/2

and 3d

5/2

electrons.

(Reproduced

from Saito, N. and Suzuki, I.H., Radiat. Res., 156, 317, 2001b.)

100 1000

W-value

Photon energy (eV)

Figure 6.16 Photon W-value for Kr as a function of photon energy. Solid circles show data by Saito and

Suzuki, and open squares are data for electrons by Combecher (1980). The solid curve represents the photon

W-values calculated using the model explained in Section 6.5.5. The bars with hatching indicate the ionization

thresholds of 3d, 3p, and 3s electrons. (Reproduced from Saito, N. and Suzuki, I.H., Radiat. Res., 156, 317,

2001b.)

126 Charged Particle and Photon Interactions with Matter

by 1 +0.75E

) + 0.25E

). In the double photoionization case, two electrons

with energies E

and E

are emitted through photoabsorption. The excess energy (E

+E

)

is calculated by assuming that two 4p electrons are ejected in double photoionization. The

distribution of excess energy into the two electrons is simply assumed so that one elec-

tron has all excess energy and the other has no energy (E

= 0). Then the total number of

electrons produced via double ionization, N

, is expressed by 2 + E

). In the triple

photoionization case, three electrons, E

, E

, and E

, are ejected through photoabsorption,

and E

are assumed to be zero. The total number of electrons, N

, is expressed as

3 + E

). The branching ratios of Kr

, Kr

, and Kr

below the 3d threshold, that is,

,R

, and R

, respectively, which can be cited from the literature (Saito and Suzuki, 1992,

1994), correspond to the branching ratios of single, double, and triple photoionizations,

respectively.

The photon W-value is derived as

R N R N R N

s d t

+ +

1 2 3

. (6.15)

Table

6.3 lists the term symbols used in the present model, along with their explanation.

(ii) Above the 3d threshold: Photoionization of a 3d electron mainly takes place (Henke etal.,

1993). The formed 3d hole is dominantly lled through the normal Auger transition, and

a double Auger decay slightly contributes to multiple ionization. The total number of elec-

trons in the normal Auger decay, N

, is given by 2 + E

) + E

), where

is the kinetic energy of Auger electrons, which is approximately calculated to be the

table 6.3

list

of t

erm

ymbols

sed

in the m

odel

roposed

for

hoton

W-

value

symbol Content

Total number of electrons through single valence ionization

Total number of electrons through double valence ionization

Total number of electrons through triple valence ionization

Total number of electrons through the normal Auger decay

Total number of electrons through the double Auger decay

Total number of electrons through the photoionization shake-off followed

the normal

Auger

decay

shdA

Total number of electrons through the photoionization shake-off followed

the double

Auger

decay

Branching ratio of single ionization

Branching ratio of double ionization

Branching ratio of triple ionization

Branching ratio of quadruple ionization

Fraction of valence i-fold ionization among the total valence ionization

Ratio of the double Auger decay to the normal Auger decay

Energy of the ith photoelectron

Energy of the normal Auger electron

dAi

Energy of the ith electron formed through the double Auger decay

Energy of the ejected electron through photoionization shake-off

shA

Energy of the normal Auger electron after the photoionization shake-off

shdA

Energy of the double Auger electron after the photoionization shake-off

New Directions in W-Value Studies 127

energy difference between the ground state of 3d hole states and that of doubly charged

states (4p

−2

) (Radzig and Smirnov, 1985; Schmidt, 1997). The total number of electrons

in the double Auger decay, N

, is given by 3 + E

) + E

dA1

), where E

dA1

denotes the kinetic energy for one of the double Auger electrons, and the energy of another

Auger electron, E

dA2

, is assumed to be zero. The energy E

dA1

is assumed to be the energy

difference between the ground state of 3d hole states and that of triply charged states. Since

normal Auger and double Auger processes produce Kr

and Kr

, the branching ratios of

these ions can be utilized for the ratios of normal to double Auger processes, which was

observed using a coincidence technique (Saito and Suzuki, 1997). We cannot neglect the

contribution of valence photoionization. The ratio of the single valence ionization corre-

sponds to the branching ratio of Kr

. We assume that the ratios of single, double, and triple

valence ionization above the 3d threshold are equal to those just below the 3d threshold,

say R

, R

, and R

, respectively. Then the ratios of the double and triple valence ioniza-

tions are calculated to be R

and R

, respectively (Saito and Suzuki, 1992,

1994).

is expressed as follows:

R N R

N R

N R R

+ + + −













+ −

1 1

1 2 1 3 1nA













. (6.16)

(iii) At energies considerably higher than the 3d ionization threshold and lower than the 3p

threshold: A photoionization shake-off process occurs, in which a 3d electron and a

valence electron are ejected at the initial photoabsorption step. The atom having two

holes turns into a triply charged ion in most cases via Auger decay and slightly into a qua-

druply charged ion via double Auger decay. The kinetic energies for the initially ejected

electrons and the Auger electron are assumed to be E

, and zero and E

shA

, respectively.

The total number of electrons via the photoionization shake-off and Auger decay (N

) is

given by 3 + E

) + E

shA

). In the case of the double Auger decay, the kinetic

energies of the two Auger electrons are assumed to be E

shdA

and zero. Then the total

number of electrons via the photoionization shake-off and double Auger decay (N

shdA

) is

given by 4 + E

) + E

shdA

). In these cases, the energies of E

, E

shA

, and

shdA

are assumed to be the differences between the corresponding ground states among

the related hole states (Radzig and Smirnov, 1985; Schmidt, 1997). The branching ratios

of normal Auger, double Auger, shake-off + Auger, and shake-off + double Auger can

be calculated using the branching ratios of Kr ions and the ratio of the double to normal

Auger processes. The ratio of the double to normal Auger transitions (R

) is cited from

the literature (Saito and Suzuki, 1997). Then W

is represented as

R N R

N R

N R R

+ + + −













+ −

1 1 1 2 1 2 1d

R N

R R

nd dA













+ − − −

























3 1 2 1

NN R N

sh shdA













(6.17)

In this equation, R

indicates the branching ratio of the double Auger decay after the

photoionization shake-off process.

128 Charged Particle and Photon Interactions with Matter

(iv) Above the 3p ionization threshold and the 3s threshold: Similar assumptions have been

made for the calculation of W

. Partial photoabsorption cross sections of 3p and 3s elec-

trons are small, and the details of the assumptions have not considerably affected the cal-

culated

values of W

Let us compare the measured data with the result calculated by this model. The curve for W

calcu-

lated in this model has been plotted as the solid line in Figure 6.16. This curve has well reproduced

the energy dependence of the experimental photon W-value. This nding suggests that the present

model seems quantitatively valid. However, a slight difference between experimental and calcu-

lated W-values is seen in the region of 125–160 eV. A possible reason may be the effect of impurity

contained in the sample gas. If this is the case, the Penning ionization may create additional ions.

This effect will lead to a lower W-value than real. However, this kind of effect is supposed to induce

lower W-values at other photon energies. Figure 6.16 essentially indicates the agreement between

the experiment and the calculation in other energy regions. Therefore, the impurity effect does

not contribute much to the discrepancy near 140 eV. Another reason may be the mixing of higher-

order radiation in the incident soft x-ray. The second-order radiation can create ions in the chamber

about two times of the rst because of similar values of W

near 140 and 280 eV. The energy of the

electron beam in the storage ring was lowered to 350MeV in the measurement of W

in this region,

and the percentage of the second-order light was supposed to be extremely low. This supposition

was conrmed from the saturation curve with an increase in the gas density in the ion chamber, as

shown in Figure 6.14. Another possibility is related to the assumption of the ratio of photoionization

shake-off in the proposed model. This shake-off process forms many electrons simultaneously, and

thus makes secondary ionizations less efcient. A smaller number of the total electrons is given by

larger ratio of this shake-off process, that is, a larger W

is derived through this assumption.

When multiply charged Kr ions collide with neutral atoms, their charge states change due to the

collision:

Kr Kr Kr Kr KE

+ +n

+ → + +

−

( )

(6.18)

→ + + −

′

−

Kr Kr e KE

( 1)+ 2+n

(6.19)

The terms KE and KE′ denote energies of exothermic and endothermic reactions. In most cases,

the total charge of the two species is taken to be the same, as given in Equation 4.18, before and

after these collisions, because the ion does not have a high kinetic energy. The collision process

expressed by Equation 6.19 takes place only with an extremely low probability. This process con-

tributes to the increase in the total charge. Therefore, this process has a slight possibility of lower-

ing the measured W

near 140 eV. This possibility is probably supported by the fact that the ground

state of a quadruply charged Kr ion is positioned at 130eV. However, the photon energy for peak-

ing the Kr

fraction in the initial photoionization step is about 290 eV and that for Kr

is about

400eV, according to the data from the literature (Saito and Suzuki, 1992). The process expressed

by Equation 6.19 should play a more signicant role at energies of about 280eV than about 140 eV.

Finally, the discrepancy near 140eV cannot be claried at present, and several factors possibly con-

tribute

to this discrepancy.

The

obtained at 1 keV in the present study, which is 27.2 ± 1.0eV, is slightly higher than

24.0 ± 0.7eV at 5.9keV that was obtained previously using the proportional counter (Borges and

Conde, 1996). Since the curve of W

in Figure 6.16 shows a tendency of decreasing with increas-

ing photon energy, this trend seems to take the present W

value close to the previous one at that

energy. However, Kr has the L-shell absorption edges around 1.7 keV, and these edges probably

yield upward jumps in the W

. If the jump is not large, energy dependence will explain the differ-

ence

of W

at the two photon energies.

New Directions in W-Value Studies 129

Figure 6.17 shows some sharp peaks of the experimental W

in the energy region for excitation of

the 3d electron into Rydberg orbitals. According to the photoabsorption spectrum (Hayaishi et al.,

1984; Saito and Suzuki, 1992), each peak corresponds to a transition of the 3d electron into a Rydberg

orbital, that is, the peak at 91.2eV comes from the transition to the 3d

5/2

−1

5p state and the peak at

92.5eV originates from transitions to the 3d

3/2

−1

5p state and to the 3d

5/2

−1

6p state. These peaks are

connected to the energy of resonant Auger electrons. Based on the studies on resonant Auger transi-

tions, the decay processes are approximately classied into two types (Carlson et al., 1989; Okada

et al., 2005). The rst type is a spectator decay, in which the excited electron remains in the orbital just

occupied, and the second is a participator. Usually the spectator decay takes place at a high probability

when the relevant orbital is a Rydberg. Then the energy of the electron formed through the spectator

is considerably lower than the valence photoelectron at the same photon energy. This lower-energy

electron induces a lower ionization yield in the ion chamber than the photoelectron. This consideration

assumes that a higher W

is obtained at the Rydberg excitation than the valence orbital ionization. The

peak width for this excitation is very narrow, and the width measured has been governed by the mono-

chromator resolution. Therefore, the maximum value of W

at the Rydberg excitation is not conclusive,

owing to the moderate resolution of the present soft x-ray monochromator. Slightly lower values for W

are derived just above the ionization thresholds. This nding is presumed to come from the emission

of 3d electrons, which have very low energies at these energies. The low energy emission serves to

produce efcient yields of ionization in the gas system.

6.5.6 ar and xe

The photon W-value of Ar is shown in the absolute scale as a function of the photon energy from

50 to 1000eV in Figure 6.18 (Saito and Suzuki, 2001a). The present data are exhibited with solid circles.

The arrow denotes the 2p electron ionization threshold. The open squares indicate the W-value for

electrons, W

, which was reported previously (Combecher, 1980). The measured W

exhibits lower

values than those for electrons below 250eV. The W

increases sharply near the 2p threshold, which

shows slightly higher or almost the same values as W

. The absolute values for W

measured here

are presumed to be correct in consideration of the agreement with the data for electrons, W

, in the

100

Photon W-value

Electron W-value

Calculated photon W-value

W-value (eV)

Photon energy (eV)

1000

Figure 6.18 Photon W-value for Ar as a function of photon energy. Solid circles show data by Saito and

Suzuki, and open squares are data for electrons by Combecher (1980). The solid curve represents the photon

W-values calculated using the model explained in Section 6.5.6. The arrow indicates the ionization threshold

of 2p electrons.

(Reproduced from Saito, N. and Suzuki, I.H., Radiat. Phys. Chem., 60, 291, 2001a.)