- Email: [email protected]
Blistering and exfoliation of 304 stainless steel studied by SEM and RBS
BLISTERING AND EXFOLIATION OF 304 STAINLESS STEEL STUDIED BY SEM AND BBS M. BRAUN, B. EMMOTH
and J.L. WHITTON*
Research Institute of Physics, Association Euratom-Studsvik, S-104 05 Stockholm, Sweden
Blistering and exfoliation induced by helium irradiation have been studied in polycrystalline stainless steel with the aim of measuring flake or blister skin thicknesses for different implantation energies in the keV region. In this work we present measurements of skin thicknesses determined both by direct SEM observations and RBS techniques. The results of the RBS studies show for all implantation energies used that the thickness measured in target atoms/unit area of a flake equals the calculated implantation range with an accuracy of about 10%. Conversely, the SEM measurements show that swelling gives rise to a geometrical skin thickness much greater than the corresponding implantation range. Thus, from the SEM and RBS data swelling as a function of the implantation energy is obtained and the linear relative swelling is shown to be strongly dependent on the energy. In addition, a comparison between the skin thickness of blisters and flakes has been made for the same material. Blisters were observed at low implantationtemperatures and room temperatures, while exfoliation occurs at elevated temperatures. The result of this comparison is that within the experimental accuracy the flake and blister thicknesses are the same.
1. Introduction Blistkring due to helium irradiation can cause serious erosion of a metal surface. The responsible processes are not fully understood, but two models are generally referred to as possible explanations. In the gas-driven model [l, 21 bubbles are assumed to grow by coalescence at a depth correlated to the projected range of the bombarding particles. Regardless of which is the mechanism for the bubble formation and growth, a high internal pressure occurs within the solid, because of the high trapping probability gf helium in many metals, at not too high temperatures. In the lateral stress model [3,4], stresses are assumed to be introduced in the implanted layer and the surface will buckle if the integrated lateral stress becomes too high. Attempts to experimentally verify various models have shown the importance of taking swelling properties into consideration [5]. Few data exist on the magnitude of this swelling, especially in the interesting region 10-100 keV, where the relative swelling varies from about 100% to a few percent. For any successful theory or model describing the blistering phenomena, swelling must be taken into account. l
H.C. @rsWl Institute, DK-2100 Copenhagen, Denmark.
Observations of the eroded surface have often been restricted to scanning electron microscopy, few other attempts of using other techniques exist [6,7]. For the experimental determination of swelling of the exfoliated layer, studies of micrographs are not sufficient and must be complemented with another independent technique such as RBS or nuclear reactions. The aim of this work is to determine the thickness and to deduce the amount of swelling by using two methods: microscopy (SEM) and scanning electron measurements of isolated surface layers with the Rutherford backscattering technique (RBS). It will be shown that a comparison between the physical thickness measured by SEM and the thickness measured in target atoms/unit area, a result of RBS, definitely indicates an absolute swelling due to the helium implantation in the whole investigated energy region, 20-80 keV. Furthermore, the linear relative swelling is shown to increase for lower energies. The swelling at lower energies 520 keV must be considered, comparably, as important an effect as blistering itself. 2. Experimental Samples of highly polished, cold-worked 304 stainless steel were irradiated with mass-analysed
Joumal of Nuclear Materials 93 & 94 (1980) 728-733 @ North-Holland Publishing Company
728
M. Braun et al. / Blistering and exfoliation of 304 SS
helium ions in the 100 keV accelerator. The vacuum in the experimental chamber was held
729
about 430K during the irradiations in these cases. The helium-bombarded targets were examined by SEM and the geometrical thicknesses of flakes or erupted blisters were measured. In the case of circular-shaped blisters it was often difficult to find erupted material to be seen in the observational direction and therefore thickness measurements in these cases had to be excluded. Finally, the erupted layers were measured with a recently developed technique [g]. After exfoliation had occurred the loosely bound top layers were isolated and fixed on a carbon tape and mounted in a target chamber for backscattering analysis. A 1.8 MeV He+ ion beam from the Van de Graaff accelerator was used in this analysis. Fig. 1 shows an example of a backscattering energy spectrum obtained from such isolated flakes on the carbon tape. It is seen that the backing material mainly consists of carbon. It was not possible to resolve Cr, Fe, Ni in this spectrum. The thickness was determined in atoms mm2 and converted to a metric scale by assuming that the density was that of bulk iron, and that the contribution of the stopping cross section from the implanted helium atoms could
Fig. 1. Typical backscattering spectrum obtained from 304 stainless steel flakes isolated on a carbon tape after 50 keV helium irradiation. The front edges of different elements in the flakes and the carbon tape are indicated. The flake thickness (in atoms m-*) is obtained from the energy difference derived from the spectrum of the stainless steel film.
730
M. Braun et al. I Blisteting and exfoliation of 304 SS
be ignored. Support for this assumption was obtained by model calculations of stopping in mixed media [9]. 3. Results Fig. 2a shows an erupted bubble after 75 keV irradiation and a dose of 7.0 x l@’ atoms m-*, slightly higher than the critical dose for blistering at this energy. Bubble formation was observed for all energies at RT irradiations. It is interesting to note that the bubble has a circular base but around its base it flakes, as we assume, along the grain boundaries. The micrograph also indicates that the thickness of the bubble is greater than the flat erupted part surrounding the bubble. At present we have no explanation for this effect which also was observed at lower energies. Since this particular bubble shown in fig. 1 has cracked across the top, it was possible to measure the blister cover thickness by SEM. It was found to be 0.31 pm. The so obtained geometrical thickness is about 15% larger than the corresponding projected range at this energy. Fig. 2b shows the result when the irradiation was done with the target held at a temperature of 410 K and 75 keV ion bombardment energy. At this temperature
large-scale exfoliation is observed all over the surface. In this case the geometrical thickness can be measured accurately from the micrographs since some flakes are twisted so that both sides of the flakes can be seen and the sight angle is parallel to the surface. It is observed from fig. 2b that the flake has a pronounced rough surface on the erupted side. This roughness can be estimated to be about 10% of the total thickness, a result, although not very accurate, that is well below the expected straggling at this energy. The thickness of this flake is found to be 0.31 pm, i.e. in close agreement with the result of the bubble thickness found for the same energy. For other energies it was not always possible to measure the skin thickness of the bubbles very accurately because of the geometrical orientation of the bubbles. An overall impression is that the bubbles in general break up outside or close to the base and seldom along its top. However, we conclude from measurements, particularly at 75 keV, that the thickness of bubbles and flakes is virtually the same, indicating that the exfoliation and bubble formation can be described with a unified model. It is only the temperature which determines if blistering or flaking takes place and thus one might only consider changes of the
Fig. 2. (a) Scanning electron micrograph of a cracked blister showing the different skin thickness of the top and the bottom of the bubble. Note the porous-like structure in the circular base of the bubble. The helium irradiation energy was 75 keV and the target temperature 310 K. (b) Detail of a twisted flake bending up from the surface. The SEM measurements are done with the help of such micrographs (see text). Note the porous structure on the ruptured side of the flake. The helium irradiation energy was 75 keV and the target temperature 430 K.
M. Braun et al. I Blistering and exfoliation of 304 SS
material properties for different temperatures under study, and not the mechanism for the surface deformation itself. Further SEM measurements were made on flakes at different energies. The geometrical flake thicknesses were measured from micrographs in the same way as indicated above and the results are plotted in fig. 3. The SEM data correspond to the * crosses. The results from the RBS measurements are also indicated in fig. 3 by the circles, as well as a comparison with projected ranges obtained from tables by Ziegler [lo] (the full-drawn line). The RBS data obtained from measurements of isolated flakes have been converted from atoms mm2 to a linear scale by assuming that the density of the flakes equals the density of the bulk material, i.e. without taking into account the swelling (see also ref. [S]). Thus, the so obtained flake thicknesses cbrrespond to values for bulk stainless steel. In contrast to the RBS data, the SEM results are independent of the composition of the material and show the geometrical thickness including swelling due to the helium implantation. All our SEM data show a discrepancy from the RBS data, which we
x SEM
T
0 RBS
731
think is due to an expansion of the material, since the SEM data of the measured thickness exceed the RBS data. It should be pointed out that, in principle, the linear expansion reported here and the volume expansion could be different. From our RBS data we find that the flakes break at a depth very close to the corresponding calculated mean projected range R,. From the comparison between the SEM and RBS data one obtains absolute values of the swelling for all energies. This absolute linear swelling (=O.O4pm) is fairly constant for all energies, probably a result of the fact that the critical dose for flaking is almost the same for all energies. However, as mentioned above, slightly higher doses were required for lower energies, but this excess of helium implantation might be compensated by the fact that part of the trapped helium is re-emitted during the irradiation for lower energies. The relative linear expansion (expressed percent-wise) as a function of the implantation energy has been plotted in fig. 4. The solid line is drawn as a guide for the eye. As a comparison to earlier data we have included results from similar measurements done with aluminium targets [8]. The relative discrepancy between SEM and RBS ._ x STAINLESS
70
STEEL
304
0 ALUMINIUM 60
‘f
\
50
LO30 20
’
IO ’ 30 ENERGY
50
20
70
IKeV)
Fig. 3. Measured skin thickness of flakes as a function helium implantation energy, determined by SEM (crosses) and RBS analyses (circles). The solid curve the calculated mean projected range given by Ziegler
LO ENERGY
of the studies shows [lo].
60 II+‘)
80
4. Relative percentage of linear expansion of exfoliated flakes as a function of helium implantation energy. The crosses refer to the results found for stainless steel in this work. As a comparison earlier data of aluminium (circles) are also included in the figure.
732
M. Braun et al. I Blistering and exfoliation of 304 SS
data should decrease for higher energies, but it is interesting to note that the linear expansion is seen to approach 70% for the lowest energies. From the data presented here it is of course difficult to predict the result for still lower energies, but one could expect that the tendency to require higher doses for lower energies is a result of increasing losses of helium during the bombardment, since it will be implanted closer to the surface, which also should limit the amount of swelling. In a comparison between swelling in stainless steel and aluminium it is found that the relative swelling is almost the same in spite of the different implantation depths of helium at these energies. The main result of this comparison is that in both cases a considerable linear expansion is found and that the ‘relative expansion increases for lower energies. It should be noted that both results were achieved under similar conditions and that the required exfoliation doses also were comparable. Different material properties could, however, result in this coincident result. 4. Conclusions Independent measurements of the flake thickness, from SEM micrographs and with RBS on isolated exfoliated target material, show a discrepancy which is explained by swelling. The energy region studied is especially interesting since in this region the relative linear expansion increases from 15% to more than 70%, while the absolute swelling is fairly constant. Only few data exist in the literature [II] but they confirm this result for the lowest energies. The influence of swelling is important and calculations performed within different models are probably much affected by this change of the material density. From a few measurements at the highest energies it was possible to compare blister skin thickness and flake thickness simply by changing the target temperature and making the irradiations under similar conditions. The result was that the skin and flake thicknesses are very close to each other, indicating that the driving mechanism is the same and that the different
kinds of erosion are due to a temperaturedependent material property. It should be noted that the measurements which have been compared in this work deal with linear expansion and this does not tell us much about the distribution of swelling. According to earlier results the most dense bubble accumulation is to be found close to the projected range of the helium ions [5,12]. However, it is still an open question from the data presented here how big the real volume expansion is [13,14]. The absolute linear swelling is found to be fairly constant as a function of energy which is also confirmed by earlier measurements on aluminium. This finding is probably connected to the very small variation in blistering dose required in this energy region as well as that the range straggling also is fairly constant. Finally, we would like to point out that if swelling due to the helium implantation is neglected and we calculate the thickness in target atoms m-*, there is good agreement between thickness and projected range in spite of the error introduced by the straggling of the helium ions. This result should be compared with the findings of ref. [7] in which niobium blisters were investigated. In ref. [7] it was found that for 30 keV the thickness measured in number of target atoms m-* with RBS double alignment techniques considerably exceeded the theoretical projected range. The advantages of the RBS method used here and applied to thickness measurements of flakes removed from the surface are (a) the measurement yields a true value of target atoms per unit area, independent of the material density, (b) the material below the ruptured region has no influence on the measurement as might be the case if the damaging effect of a single crystal is determined by channeling techniques, and (c) both polycrystalline and single crystal specimens can be used. Acknowledgement The authors wish to express their gratitude to the National Swedish Board for Energy Source Development who supported this work financially.
M. Braun et al. / Blistering and exfoliationof 304 SS
References [l] SK. Das, M. Kaminsky and Cl. Fenske, J. Nucl. Mater. 76/77 (1978) 215. [2] S.K. Das, M. Kaminsky and G. Fenske, J. Appl. Phys. 50 (1979) 3304. [3] J. Roth, in: Application of Ion Beams to Materials, Eds. G. Carter, J.S. Colligan and N.A. Grant (The Institute of Physics, London, 1976) p. 280. [4] E.P. EerNisse and ST. Picraux, J. Appl. Phys. 48 (1977) [5] ‘d. Fenske, S.K. Das, M. Kaminsky and G.H. Miley, J. Nucl. Mater. 76/77 (1978) 247. [6] R. Behrisch, J. Bettiger, W. Eckstein, U. Littmark, J. Roth and B.M.U. Scherzer, Appl. Phys. Letters 27 (1975) 199.
733
[7] M.R. Risch, J. Roth and B.M.U. Scherzer, J. Nucl. Mater. 82 (1979) 220. [8] M. Braun, J.L. Whitton and B. Emmoth, J. Nucl. Mater. 85/86 (1979) 1091. [9] Ion Beam Handbook, Eds. Mayer and Rimini (Academic Press, 1977). [lo] Stopping Powers and Ranges in All Elemental Matter, Ed. J.F. Ziegler (Pergamon, 1977). [ll] R.G. St-Jacques, G. Veilleux, J.G. Martel and B. Terresult, Radiation Effects, to be published. [12] S.E. Donelly, G. Debras, J.-M. Gilles and A.A. Lucas, Radiation Effects Letters 50 (1980) 57. [13] R.S. Blewer and W. Beezhold, Radiation Effects 19 (1973) 49. [14] J.H. Evans, J. Nucl. Mater. 76/77 (1978) 228.
and J.L. WHITTON*
Research Institute of Physics, Association Euratom-Studsvik, S-104 05 Stockholm, Sweden
Blistering and exfoliation induced by helium irradiation have been studied in polycrystalline stainless steel with the aim of measuring flake or blister skin thicknesses for different implantation energies in the keV region. In this work we present measurements of skin thicknesses determined both by direct SEM observations and RBS techniques. The results of the RBS studies show for all implantation energies used that the thickness measured in target atoms/unit area of a flake equals the calculated implantation range with an accuracy of about 10%. Conversely, the SEM measurements show that swelling gives rise to a geometrical skin thickness much greater than the corresponding implantation range. Thus, from the SEM and RBS data swelling as a function of the implantation energy is obtained and the linear relative swelling is shown to be strongly dependent on the energy. In addition, a comparison between the skin thickness of blisters and flakes has been made for the same material. Blisters were observed at low implantationtemperatures and room temperatures, while exfoliation occurs at elevated temperatures. The result of this comparison is that within the experimental accuracy the flake and blister thicknesses are the same.
1. Introduction Blistkring due to helium irradiation can cause serious erosion of a metal surface. The responsible processes are not fully understood, but two models are generally referred to as possible explanations. In the gas-driven model [l, 21 bubbles are assumed to grow by coalescence at a depth correlated to the projected range of the bombarding particles. Regardless of which is the mechanism for the bubble formation and growth, a high internal pressure occurs within the solid, because of the high trapping probability gf helium in many metals, at not too high temperatures. In the lateral stress model [3,4], stresses are assumed to be introduced in the implanted layer and the surface will buckle if the integrated lateral stress becomes too high. Attempts to experimentally verify various models have shown the importance of taking swelling properties into consideration [5]. Few data exist on the magnitude of this swelling, especially in the interesting region 10-100 keV, where the relative swelling varies from about 100% to a few percent. For any successful theory or model describing the blistering phenomena, swelling must be taken into account. l
H.C. @rsWl Institute, DK-2100 Copenhagen, Denmark.
Observations of the eroded surface have often been restricted to scanning electron microscopy, few other attempts of using other techniques exist [6,7]. For the experimental determination of swelling of the exfoliated layer, studies of micrographs are not sufficient and must be complemented with another independent technique such as RBS or nuclear reactions. The aim of this work is to determine the thickness and to deduce the amount of swelling by using two methods: microscopy (SEM) and scanning electron measurements of isolated surface layers with the Rutherford backscattering technique (RBS). It will be shown that a comparison between the physical thickness measured by SEM and the thickness measured in target atoms/unit area, a result of RBS, definitely indicates an absolute swelling due to the helium implantation in the whole investigated energy region, 20-80 keV. Furthermore, the linear relative swelling is shown to increase for lower energies. The swelling at lower energies 520 keV must be considered, comparably, as important an effect as blistering itself. 2. Experimental Samples of highly polished, cold-worked 304 stainless steel were irradiated with mass-analysed
Joumal of Nuclear Materials 93 & 94 (1980) 728-733 @ North-Holland Publishing Company
728
M. Braun et al. / Blistering and exfoliation of 304 SS
helium ions in the 100 keV accelerator. The vacuum in the experimental chamber was held
729
about 430K during the irradiations in these cases. The helium-bombarded targets were examined by SEM and the geometrical thicknesses of flakes or erupted blisters were measured. In the case of circular-shaped blisters it was often difficult to find erupted material to be seen in the observational direction and therefore thickness measurements in these cases had to be excluded. Finally, the erupted layers were measured with a recently developed technique [g]. After exfoliation had occurred the loosely bound top layers were isolated and fixed on a carbon tape and mounted in a target chamber for backscattering analysis. A 1.8 MeV He+ ion beam from the Van de Graaff accelerator was used in this analysis. Fig. 1 shows an example of a backscattering energy spectrum obtained from such isolated flakes on the carbon tape. It is seen that the backing material mainly consists of carbon. It was not possible to resolve Cr, Fe, Ni in this spectrum. The thickness was determined in atoms mm2 and converted to a metric scale by assuming that the density was that of bulk iron, and that the contribution of the stopping cross section from the implanted helium atoms could
Fig. 1. Typical backscattering spectrum obtained from 304 stainless steel flakes isolated on a carbon tape after 50 keV helium irradiation. The front edges of different elements in the flakes and the carbon tape are indicated. The flake thickness (in atoms m-*) is obtained from the energy difference derived from the spectrum of the stainless steel film.
730
M. Braun et al. I Blisteting and exfoliation of 304 SS
be ignored. Support for this assumption was obtained by model calculations of stopping in mixed media [9]. 3. Results Fig. 2a shows an erupted bubble after 75 keV irradiation and a dose of 7.0 x l@’ atoms m-*, slightly higher than the critical dose for blistering at this energy. Bubble formation was observed for all energies at RT irradiations. It is interesting to note that the bubble has a circular base but around its base it flakes, as we assume, along the grain boundaries. The micrograph also indicates that the thickness of the bubble is greater than the flat erupted part surrounding the bubble. At present we have no explanation for this effect which also was observed at lower energies. Since this particular bubble shown in fig. 1 has cracked across the top, it was possible to measure the blister cover thickness by SEM. It was found to be 0.31 pm. The so obtained geometrical thickness is about 15% larger than the corresponding projected range at this energy. Fig. 2b shows the result when the irradiation was done with the target held at a temperature of 410 K and 75 keV ion bombardment energy. At this temperature
large-scale exfoliation is observed all over the surface. In this case the geometrical thickness can be measured accurately from the micrographs since some flakes are twisted so that both sides of the flakes can be seen and the sight angle is parallel to the surface. It is observed from fig. 2b that the flake has a pronounced rough surface on the erupted side. This roughness can be estimated to be about 10% of the total thickness, a result, although not very accurate, that is well below the expected straggling at this energy. The thickness of this flake is found to be 0.31 pm, i.e. in close agreement with the result of the bubble thickness found for the same energy. For other energies it was not always possible to measure the skin thickness of the bubbles very accurately because of the geometrical orientation of the bubbles. An overall impression is that the bubbles in general break up outside or close to the base and seldom along its top. However, we conclude from measurements, particularly at 75 keV, that the thickness of bubbles and flakes is virtually the same, indicating that the exfoliation and bubble formation can be described with a unified model. It is only the temperature which determines if blistering or flaking takes place and thus one might only consider changes of the
Fig. 2. (a) Scanning electron micrograph of a cracked blister showing the different skin thickness of the top and the bottom of the bubble. Note the porous-like structure in the circular base of the bubble. The helium irradiation energy was 75 keV and the target temperature 310 K. (b) Detail of a twisted flake bending up from the surface. The SEM measurements are done with the help of such micrographs (see text). Note the porous structure on the ruptured side of the flake. The helium irradiation energy was 75 keV and the target temperature 430 K.
M. Braun et al. I Blistering and exfoliation of 304 SS
material properties for different temperatures under study, and not the mechanism for the surface deformation itself. Further SEM measurements were made on flakes at different energies. The geometrical flake thicknesses were measured from micrographs in the same way as indicated above and the results are plotted in fig. 3. The SEM data correspond to the * crosses. The results from the RBS measurements are also indicated in fig. 3 by the circles, as well as a comparison with projected ranges obtained from tables by Ziegler [lo] (the full-drawn line). The RBS data obtained from measurements of isolated flakes have been converted from atoms mm2 to a linear scale by assuming that the density of the flakes equals the density of the bulk material, i.e. without taking into account the swelling (see also ref. [S]). Thus, the so obtained flake thicknesses cbrrespond to values for bulk stainless steel. In contrast to the RBS data, the SEM results are independent of the composition of the material and show the geometrical thickness including swelling due to the helium implantation. All our SEM data show a discrepancy from the RBS data, which we
x SEM
T
0 RBS
731
think is due to an expansion of the material, since the SEM data of the measured thickness exceed the RBS data. It should be pointed out that, in principle, the linear expansion reported here and the volume expansion could be different. From our RBS data we find that the flakes break at a depth very close to the corresponding calculated mean projected range R,. From the comparison between the SEM and RBS data one obtains absolute values of the swelling for all energies. This absolute linear swelling (=O.O4pm) is fairly constant for all energies, probably a result of the fact that the critical dose for flaking is almost the same for all energies. However, as mentioned above, slightly higher doses were required for lower energies, but this excess of helium implantation might be compensated by the fact that part of the trapped helium is re-emitted during the irradiation for lower energies. The relative linear expansion (expressed percent-wise) as a function of the implantation energy has been plotted in fig. 4. The solid line is drawn as a guide for the eye. As a comparison to earlier data we have included results from similar measurements done with aluminium targets [8]. The relative discrepancy between SEM and RBS ._ x STAINLESS
70
STEEL
304
0 ALUMINIUM 60
‘f
\
50
LO30 20
’
IO ’ 30 ENERGY
50
20
70
IKeV)
Fig. 3. Measured skin thickness of flakes as a function helium implantation energy, determined by SEM (crosses) and RBS analyses (circles). The solid curve the calculated mean projected range given by Ziegler
LO ENERGY
of the studies shows [lo].
60 II+‘)
80
4. Relative percentage of linear expansion of exfoliated flakes as a function of helium implantation energy. The crosses refer to the results found for stainless steel in this work. As a comparison earlier data of aluminium (circles) are also included in the figure.
732
M. Braun et al. I Blistering and exfoliation of 304 SS
data should decrease for higher energies, but it is interesting to note that the linear expansion is seen to approach 70% for the lowest energies. From the data presented here it is of course difficult to predict the result for still lower energies, but one could expect that the tendency to require higher doses for lower energies is a result of increasing losses of helium during the bombardment, since it will be implanted closer to the surface, which also should limit the amount of swelling. In a comparison between swelling in stainless steel and aluminium it is found that the relative swelling is almost the same in spite of the different implantation depths of helium at these energies. The main result of this comparison is that in both cases a considerable linear expansion is found and that the ‘relative expansion increases for lower energies. It should be noted that both results were achieved under similar conditions and that the required exfoliation doses also were comparable. Different material properties could, however, result in this coincident result. 4. Conclusions Independent measurements of the flake thickness, from SEM micrographs and with RBS on isolated exfoliated target material, show a discrepancy which is explained by swelling. The energy region studied is especially interesting since in this region the relative linear expansion increases from 15% to more than 70%, while the absolute swelling is fairly constant. Only few data exist in the literature [II] but they confirm this result for the lowest energies. The influence of swelling is important and calculations performed within different models are probably much affected by this change of the material density. From a few measurements at the highest energies it was possible to compare blister skin thickness and flake thickness simply by changing the target temperature and making the irradiations under similar conditions. The result was that the skin and flake thicknesses are very close to each other, indicating that the driving mechanism is the same and that the different
kinds of erosion are due to a temperaturedependent material property. It should be noted that the measurements which have been compared in this work deal with linear expansion and this does not tell us much about the distribution of swelling. According to earlier results the most dense bubble accumulation is to be found close to the projected range of the helium ions [5,12]. However, it is still an open question from the data presented here how big the real volume expansion is [13,14]. The absolute linear swelling is found to be fairly constant as a function of energy which is also confirmed by earlier measurements on aluminium. This finding is probably connected to the very small variation in blistering dose required in this energy region as well as that the range straggling also is fairly constant. Finally, we would like to point out that if swelling due to the helium implantation is neglected and we calculate the thickness in target atoms m-*, there is good agreement between thickness and projected range in spite of the error introduced by the straggling of the helium ions. This result should be compared with the findings of ref. [7] in which niobium blisters were investigated. In ref. [7] it was found that for 30 keV the thickness measured in number of target atoms m-* with RBS double alignment techniques considerably exceeded the theoretical projected range. The advantages of the RBS method used here and applied to thickness measurements of flakes removed from the surface are (a) the measurement yields a true value of target atoms per unit area, independent of the material density, (b) the material below the ruptured region has no influence on the measurement as might be the case if the damaging effect of a single crystal is determined by channeling techniques, and (c) both polycrystalline and single crystal specimens can be used. Acknowledgement The authors wish to express their gratitude to the National Swedish Board for Energy Source Development who supported this work financially.
M. Braun et al. / Blistering and exfoliationof 304 SS
References [l] SK. Das, M. Kaminsky and Cl. Fenske, J. Nucl. Mater. 76/77 (1978) 215. [2] S.K. Das, M. Kaminsky and G. Fenske, J. Appl. Phys. 50 (1979) 3304. [3] J. Roth, in: Application of Ion Beams to Materials, Eds. G. Carter, J.S. Colligan and N.A. Grant (The Institute of Physics, London, 1976) p. 280. [4] E.P. EerNisse and ST. Picraux, J. Appl. Phys. 48 (1977) [5] ‘d. Fenske, S.K. Das, M. Kaminsky and G.H. Miley, J. Nucl. Mater. 76/77 (1978) 247. [6] R. Behrisch, J. Bettiger, W. Eckstein, U. Littmark, J. Roth and B.M.U. Scherzer, Appl. Phys. Letters 27 (1975) 199.
733
[7] M.R. Risch, J. Roth and B.M.U. Scherzer, J. Nucl. Mater. 82 (1979) 220. [8] M. Braun, J.L. Whitton and B. Emmoth, J. Nucl. Mater. 85/86 (1979) 1091. [9] Ion Beam Handbook, Eds. Mayer and Rimini (Academic Press, 1977). [lo] Stopping Powers and Ranges in All Elemental Matter, Ed. J.F. Ziegler (Pergamon, 1977). [ll] R.G. St-Jacques, G. Veilleux, J.G. Martel and B. Terresult, Radiation Effects, to be published. [12] S.E. Donelly, G. Debras, J.-M. Gilles and A.A. Lucas, Radiation Effects Letters 50 (1980) 57. [13] R.S. Blewer and W. Beezhold, Radiation Effects 19 (1973) 49. [14] J.H. Evans, J. Nucl. Mater. 76/77 (1978) 228.