- Email: [email protected]
Interaction of a two-component hydrogen-containing glow discharge plasma with Pd and stainless steel
Vacuum/volume
43lnumbers Printed in Great Britain
5-7lpages
0042-207x/92$5.00+.00 @ 1992 Pergamon Press Ltd
623 to 625/I 992
Interaction of a two-component hydrogencontaining glow discharge plasma with Pd and stainless steel A V Sharudo Uzbekistan
and R E Mukhamadiev,
Arifov Institute
of Electronics,
Academgorodok,
Tashkent
700743,
The interaction of a two-component hydrogen-containing glow discharge plasma (q = 3~ 7016 ion crnd2 s-‘, U = 250-450 V) with Ar, He, N2 and O2 impurities (Pi,,,,,/?, = O-7.0) with Pd and stainless steel has been studied using TPD, DRS and hydrogen permeability techniques. It was shown that for Pd the rate limiting step for hydrogen penetration is surface processes while for stainless steel the penetration is limited by bulk diffusion. Activation of penetration processes by glow discharge plasma causes an increase in the penetrating flow from 4.2~ 7014 to 5.6~ 7018 atom cmm2 s-’ (T,,,,, = 400 K) and from 7.2 x 7015 to 4.7 x 70’” atom cmm2 s-’ (T,,, = 470 K), respectively. The addition of impurities into the hydrogen plasma results in a decrease of the penetrating flow by a factor of 2-3 for inert gases and more than an order of magnitude for impurities of chemically active gases. Changes in the hydrogen re-emission rate constant. SR, depending on plasma composition was shown for stainless steel surfaces. The re-emission rate constant ranged from 7.58 to 2.63~ 70-2 cm s-’ for He impurities, from 7.58 to 3.05~ 70m2 cm s-’ for Ar impurities and from 7.58 to 70.5~ 70-2 cm s-’ for N2 impurities.
1. Introduction Active progress of hydrogen power-engineering and vacuum engineering has raised interest in hydrogen interaction with metals, in particular, in the fields such as hydrogenation of structural elements, their embrittlement (or embrittlement resistance). During recent years development of ecologically unpolluted technologies with using hydrogen fuel gains importance. Taking into account the significance of the above-mentioned problems, the aim of the present paper is to study interaction of a two-component hydrogen-containing glow discharge plasma (U = 25045OV, 4 = 3 x 10’6ioncm~2s~ ‘) with Pd and stainless steel. Inert (He, Ar) or chemically active (NZ, 0,) gases with a ratio relative to hydrogen of P,,,/P, = O-1 .O (PH = const) were used as impurities. The analysis of the energy distribution of positive ions in such a plasma over the pressure range of 1.3-6.65 x IO2 Pa showed that the maximum of ion flow in all cases was between 70 and 90 eV. This confirms multiple charge exchange of ions moving towards the cathode. Charge exchange varied from 9 to I3 in number for different gases. Taking energy distribution into account, the process of hydrogen penetration through the samples can be described by a usual diffusion equation with the implanted ion range distribution as the corresponding source function :
implantation depth (R,) and by the concentration depth gradient. It was also found that the ionized hydrogen component contributes slightly in the penetrating flow - (lo- 3-10m “) x J. To visualize the hydrogen penetration process, the balance equation of input (Q), inverse re-emission (F) and penetrating (J) flows was assumed i.e. Q = F+J. These flows are described in terms of re-emission (S,) and thermodesorption (S,) rate constants, diffusion coefficient (II) and gas concentration on the surface (C,, CJ :
J=W-G)
s
-
1
(2)
In this case SR = &+S,, where S, is the rate constant of ioninduced desorption. Expressions for J and Fin the given form are applicable if RP - AR, which was the case for the used ion energies. Depending on the ratio of values S, and D/l the next particular cases are possible : (i) penetration limitation by diffusion (S, >> D/l)
Q/J=
l+;;; limited by surface processes
(S, << O/l)
C(X=O)=C,
Ddjj’ +NX) =O, q-y= where N(X) = q/R, energy range (under surface ; q is the ion depth. The solution flow is independent determined by the
(1)
2,
F = &C,.
(ii) penetration
d2C
c
T
,)
=
c2
exp (--X/R,) is applicable to the whole I keV) at the normal ion incidence on the flow density ; R, is the average implantation of this equation has shown that penetrating of the ion range distribution function and is membrane thickness (I), the average ion
Q/J=
l+$
Hydrogen temperature membrane membrane is observed
r
permeability studies under thermal activation in range of 32&570 K have shown that : (i) for a the value of the penetrating flow is independent thickness while for stainless steel a dependence J -
(4) the Pd of I- ’
: (ii) there are two inflections on the In J = f’(l /T) 623
A V Sharudo
and R E Mukhamadiev:
Glow discharge
plasma
08
Figure 1. Temperature dependence of penetrating hydrogen flow through 50 (0) and 100 (A) pm thick Pd under thermal activation (1, 2) and glow discharge plasma (3. 4) conditions.
curve for the Pd membrane at the temperatures of 383 and 423 K. At these points, changes in the penetration activation energies were observed (Figure 1) ; and (iii) for stainless steel membranes and a temperature range of 370-570 K, the penetration activation with the diffusion activation energy energy E,,, coincides Edll. = 6.3kO.3 Kcal mol- ‘. To explain the appearance of inflections in the In ./ = f’( l/T) curves for Pd membranes, we have carried out investigations on hydrogen thermodesorption from Pd using the thermoprogrammed desorption technique (TPD). Exposure in hydrogen was done at room temperature and PH = lo5 Pa. Two peaks were found at 383 and 423 K, which, in our opinion, are due to various sorption centres on the Pd surface. This was confirmed by the studies performed with diffusion reflection spectroscopy (DRS) which revealed the existence of Pd ions on the surface in various valence states, Pd(l), Pd( 1l), Pd(l11). The existing different sorption centres were identified with Pd ions in various valence states. Evaluation of the thermodesorption activation energy for the above-mentioned peaks yields : E = 11.7 k 0.3 Kcal mol- ’ for T = 383 K and E = 23.7 kO.3 Kcal mol ’ for T = 423 K. A comparison between the obtained thermodesorption activation energies and hydrogen penetration activation energies has shown that they completely correlate with each other. During ignition of a glow discharge plasma the penetrating flow increased for a Pd membrane from 4.2 x lOI to 5.6 x 10” atoms cm ’ s ’ (T,,,, = 400 K), and for a stainless steel membrane from 1.2x 10” to 4.1 x IO” atoms cm-‘s-’ (T,,,, = 470 K). The dependence of J and I was the same as in the case of thermal activation. These data allow the conclusion that for Pd membranes the limiting stages of penetration are the processes on the surface and for stainless steel membranes the hydrogen bulk diffusion. In accordance with this, changes in J for the said membranes can be described by equations (4) and (3), respectively. Proceeding from this fact, on the basis of equations (4) and experimental results for Pd membranes, we have determined the ratio of thermodesorption (S,) and ion-induced desorption (S,) rate constants. Out of the data presented in Figure 1 it follows that providing only the ionized component q as a flow onto the surface-the ratio S,/& turns out to be less than zero which does not make sense. To eliminate this contradiction. the 624
IO
Ratio t&/P,) Figure 2. Change of the rate constant of hydrogen re-emission S, from stainless steel surface under two-component plasma conditions : (I) hydrogen-helium : (2) hydrogen-argon : and (3) hydrogen~ nitrogen.
dependence of J on membrane polarity was studied and it was shown that within experimental limits J is independent of membrane polarity. This indicates that the ionized hydrogen component has only little effect on the flow (J), i.e. S, CCS,-. Taking this into account, evaluation of flow (Q) from equation (4) gives q << Q = 1.1 x 10 ’ ’ atoms cm-’ s ‘. The analysis of In J =.f’(l/7J for a Pd membrane under glow discharge plasma shows that this dependence corresponds to hydrogen penetration from an atomic phase’. Hence, the value obtained of the incoming flow (Q) is conditioned by the flow of the atomic component onto the surface. After determination of the value of real flow (Q) on the membrane surface using equation (3) it is possible to determine the hydrogen t-e-emission rate for stainless steel surfaces. The measured value (S,) turned out to be 1.58 x IO ’ cm s ‘_ Due to the fact that hydrogen penetration is limited by bulk diffusion (E,,,, = I!?,,,,),the re-emission rate is temperature independent. Furthermore membrane permeability has been studied under two-component plasma action as function of partial composition. In all cases a reduction of J was observed with an increasing impurity content. Inert gas impurities lower J by a factor of 2-3 and chemically active gas impurities yield a reduction of an order of magnitude. The observable change of J is related to a variation of the ratio between input, penetrating and re-emission flows which is conditioned by plasma~
A V Sharudo
and R E Mukhamadiev:
Glow discharge
plasma
2. Conclusions
Our results can be summarised
as follows.
(1) It is shown that the penetrating gas flow through metallic membranes under ion implantation of 1 keV particles is independent of the ion range distribution and is defined by membrane thickness, average implantation depth and concentration gradient. (2) It is established that under glow discharge plasma conditions (U = 25&45OV, q = 3 x lO’“ioncm~‘s~ ‘) the hydrogen penetration through Pd membranes is limited by surface processes and that the ion-induced desorption rate constant S, CCST. (3) It is shown that hydrogen penetration through stainless steel at 37C-570 K and irradiation in plasma (U = 25&450 V, q=3xlO’h atoms cmm2 SC’) is limited by bulk diffusion. It
is shown that the hydrogen re-emission rate constant (S,) is independent of temperature (T = 37&570 K) and is 1.58 x lo- * cm ss’. (4) In two-component hydrogen plasmas with P,,,/P, =& 1.0 an increase is observed in the rate constant of hydrogen re-emission from the stainless steel surface. For Ar impurities this increase was 1.58-3.02 x lo- * cm so ‘, for He impurities 1.58-2.63 x lo-’ cm ss’ and for Nz impurities 1.58-10.5 x 1O-2 cm s- ‘. References
’ A Yu Doroshin, A A Samartsev and A I Livshits, Surface, 8, 31 (1985) (in Russian). ‘1 Rot, Chemical Sputtering of So1id.s Under Ion Bombardment. MIR, Moscow (1986) (in Russian).
625
43lnumbers Printed in Great Britain
5-7lpages
0042-207x/92$5.00+.00 @ 1992 Pergamon Press Ltd
623 to 625/I 992
Interaction of a two-component hydrogencontaining glow discharge plasma with Pd and stainless steel A V Sharudo Uzbekistan
and R E Mukhamadiev,
Arifov Institute
of Electronics,
Academgorodok,
Tashkent
700743,
The interaction of a two-component hydrogen-containing glow discharge plasma (q = 3~ 7016 ion crnd2 s-‘, U = 250-450 V) with Ar, He, N2 and O2 impurities (Pi,,,,,/?, = O-7.0) with Pd and stainless steel has been studied using TPD, DRS and hydrogen permeability techniques. It was shown that for Pd the rate limiting step for hydrogen penetration is surface processes while for stainless steel the penetration is limited by bulk diffusion. Activation of penetration processes by glow discharge plasma causes an increase in the penetrating flow from 4.2~ 7014 to 5.6~ 7018 atom cmm2 s-’ (T,,,,, = 400 K) and from 7.2 x 7015 to 4.7 x 70’” atom cmm2 s-’ (T,,, = 470 K), respectively. The addition of impurities into the hydrogen plasma results in a decrease of the penetrating flow by a factor of 2-3 for inert gases and more than an order of magnitude for impurities of chemically active gases. Changes in the hydrogen re-emission rate constant. SR, depending on plasma composition was shown for stainless steel surfaces. The re-emission rate constant ranged from 7.58 to 2.63~ 70-2 cm s-’ for He impurities, from 7.58 to 3.05~ 70m2 cm s-’ for Ar impurities and from 7.58 to 70.5~ 70-2 cm s-’ for N2 impurities.
1. Introduction Active progress of hydrogen power-engineering and vacuum engineering has raised interest in hydrogen interaction with metals, in particular, in the fields such as hydrogenation of structural elements, their embrittlement (or embrittlement resistance). During recent years development of ecologically unpolluted technologies with using hydrogen fuel gains importance. Taking into account the significance of the above-mentioned problems, the aim of the present paper is to study interaction of a two-component hydrogen-containing glow discharge plasma (U = 25045OV, 4 = 3 x 10’6ioncm~2s~ ‘) with Pd and stainless steel. Inert (He, Ar) or chemically active (NZ, 0,) gases with a ratio relative to hydrogen of P,,,/P, = O-1 .O (PH = const) were used as impurities. The analysis of the energy distribution of positive ions in such a plasma over the pressure range of 1.3-6.65 x IO2 Pa showed that the maximum of ion flow in all cases was between 70 and 90 eV. This confirms multiple charge exchange of ions moving towards the cathode. Charge exchange varied from 9 to I3 in number for different gases. Taking energy distribution into account, the process of hydrogen penetration through the samples can be described by a usual diffusion equation with the implanted ion range distribution as the corresponding source function :
implantation depth (R,) and by the concentration depth gradient. It was also found that the ionized hydrogen component contributes slightly in the penetrating flow - (lo- 3-10m “) x J. To visualize the hydrogen penetration process, the balance equation of input (Q), inverse re-emission (F) and penetrating (J) flows was assumed i.e. Q = F+J. These flows are described in terms of re-emission (S,) and thermodesorption (S,) rate constants, diffusion coefficient (II) and gas concentration on the surface (C,, CJ :
J=W-G)
s
-
1
(2)
In this case SR = &+S,, where S, is the rate constant of ioninduced desorption. Expressions for J and Fin the given form are applicable if RP - AR, which was the case for the used ion energies. Depending on the ratio of values S, and D/l the next particular cases are possible : (i) penetration limitation by diffusion (S, >> D/l)
Q/J=
l+;;; limited by surface processes
(S, << O/l)
C(X=O)=C,
Ddjj’ +NX) =O, q-y= where N(X) = q/R, energy range (under surface ; q is the ion depth. The solution flow is independent determined by the
(1)
2,
F = &C,.
(ii) penetration
d2C
c
T
,)
=
c2
exp (--X/R,) is applicable to the whole I keV) at the normal ion incidence on the flow density ; R, is the average implantation of this equation has shown that penetrating of the ion range distribution function and is membrane thickness (I), the average ion
Q/J=
l+$
Hydrogen temperature membrane membrane is observed
r
permeability studies under thermal activation in range of 32&570 K have shown that : (i) for a the value of the penetrating flow is independent thickness while for stainless steel a dependence J -
(4) the Pd of I- ’
: (ii) there are two inflections on the In J = f’(l /T) 623
A V Sharudo
and R E Mukhamadiev:
Glow discharge
plasma
08
Figure 1. Temperature dependence of penetrating hydrogen flow through 50 (0) and 100 (A) pm thick Pd under thermal activation (1, 2) and glow discharge plasma (3. 4) conditions.
curve for the Pd membrane at the temperatures of 383 and 423 K. At these points, changes in the penetration activation energies were observed (Figure 1) ; and (iii) for stainless steel membranes and a temperature range of 370-570 K, the penetration activation with the diffusion activation energy energy E,,, coincides Edll. = 6.3kO.3 Kcal mol- ‘. To explain the appearance of inflections in the In ./ = f’( l/T) curves for Pd membranes, we have carried out investigations on hydrogen thermodesorption from Pd using the thermoprogrammed desorption technique (TPD). Exposure in hydrogen was done at room temperature and PH = lo5 Pa. Two peaks were found at 383 and 423 K, which, in our opinion, are due to various sorption centres on the Pd surface. This was confirmed by the studies performed with diffusion reflection spectroscopy (DRS) which revealed the existence of Pd ions on the surface in various valence states, Pd(l), Pd( 1l), Pd(l11). The existing different sorption centres were identified with Pd ions in various valence states. Evaluation of the thermodesorption activation energy for the above-mentioned peaks yields : E = 11.7 k 0.3 Kcal mol- ’ for T = 383 K and E = 23.7 kO.3 Kcal mol ’ for T = 423 K. A comparison between the obtained thermodesorption activation energies and hydrogen penetration activation energies has shown that they completely correlate with each other. During ignition of a glow discharge plasma the penetrating flow increased for a Pd membrane from 4.2 x lOI to 5.6 x 10” atoms cm ’ s ’ (T,,,, = 400 K), and for a stainless steel membrane from 1.2x 10” to 4.1 x IO” atoms cm-‘s-’ (T,,,, = 470 K). The dependence of J and I was the same as in the case of thermal activation. These data allow the conclusion that for Pd membranes the limiting stages of penetration are the processes on the surface and for stainless steel membranes the hydrogen bulk diffusion. In accordance with this, changes in J for the said membranes can be described by equations (4) and (3), respectively. Proceeding from this fact, on the basis of equations (4) and experimental results for Pd membranes, we have determined the ratio of thermodesorption (S,) and ion-induced desorption (S,) rate constants. Out of the data presented in Figure 1 it follows that providing only the ionized component q as a flow onto the surface-the ratio S,/& turns out to be less than zero which does not make sense. To eliminate this contradiction. the 624
IO
Ratio t&/P,) Figure 2. Change of the rate constant of hydrogen re-emission S, from stainless steel surface under two-component plasma conditions : (I) hydrogen-helium : (2) hydrogen-argon : and (3) hydrogen~ nitrogen.
dependence of J on membrane polarity was studied and it was shown that within experimental limits J is independent of membrane polarity. This indicates that the ionized hydrogen component has only little effect on the flow (J), i.e. S, CCS,-. Taking this into account, evaluation of flow (Q) from equation (4) gives q << Q = 1.1 x 10 ’ ’ atoms cm-’ s ‘. The analysis of In J =.f’(l/7J for a Pd membrane under glow discharge plasma shows that this dependence corresponds to hydrogen penetration from an atomic phase’. Hence, the value obtained of the incoming flow (Q) is conditioned by the flow of the atomic component onto the surface. After determination of the value of real flow (Q) on the membrane surface using equation (3) it is possible to determine the hydrogen t-e-emission rate for stainless steel surfaces. The measured value (S,) turned out to be 1.58 x IO ’ cm s ‘_ Due to the fact that hydrogen penetration is limited by bulk diffusion (E,,,, = I!?,,,,),the re-emission rate is temperature independent. Furthermore membrane permeability has been studied under two-component plasma action as function of partial composition. In all cases a reduction of J was observed with an increasing impurity content. Inert gas impurities lower J by a factor of 2-3 and chemically active gas impurities yield a reduction of an order of magnitude. The observable change of J is related to a variation of the ratio between input, penetrating and re-emission flows which is conditioned by plasma~
A V Sharudo
and R E Mukhamadiev:
Glow discharge
plasma
2. Conclusions
Our results can be summarised
as follows.
(1) It is shown that the penetrating gas flow through metallic membranes under ion implantation of 1 keV particles is independent of the ion range distribution and is defined by membrane thickness, average implantation depth and concentration gradient. (2) It is established that under glow discharge plasma conditions (U = 25&45OV, q = 3 x lO’“ioncm~‘s~ ‘) the hydrogen penetration through Pd membranes is limited by surface processes and that the ion-induced desorption rate constant S, CCST. (3) It is shown that hydrogen penetration through stainless steel at 37C-570 K and irradiation in plasma (U = 25&450 V, q=3xlO’h atoms cmm2 SC’) is limited by bulk diffusion. It
is shown that the hydrogen re-emission rate constant (S,) is independent of temperature (T = 37&570 K) and is 1.58 x lo- * cm ss’. (4) In two-component hydrogen plasmas with P,,,/P, =& 1.0 an increase is observed in the rate constant of hydrogen re-emission from the stainless steel surface. For Ar impurities this increase was 1.58-3.02 x lo- * cm so ‘, for He impurities 1.58-2.63 x lo-’ cm ss’ and for Nz impurities 1.58-10.5 x 1O-2 cm s- ‘. References
’ A Yu Doroshin, A A Samartsev and A I Livshits, Surface, 8, 31 (1985) (in Russian). ‘1 Rot, Chemical Sputtering of So1id.s Under Ion Bombardment. MIR, Moscow (1986) (in Russian).
625