- Email: [email protected]
Development of a column delivering a collimated stream of Cs0 for SIMS purposes
Applied Surface Science 231–232 (2004) 940–944
Development of a column delivering a collimated stream of Cs0 for SIMS purposes T. Wirtz*, H.-N. Migeon Laboratoire d’Analyse des Mate´riaux, Centre de Recherche Public—Gabriel Lippmann, 162A avenue de la Faı¨encerie, L-1511 Luxembourg, Luxembourg Available online 30 April 2004
Abstract In order to optimize SIMS analyses using the presence of reactive Cs for the purpose of enhancing negative secondary ion emission or working in the MCsxþ mode, we have developed a column that delivers a collimated and adjustable stream of neutral Cs atoms to be deposited on the surface of the sample while this one is being analyzed. Using this new column, it was possible to introduce an analysis technique consisting of a Xyþ ion bombardment (where X stands for any element excepting Cs) accompanied by a simultaneous deposition of Cs0 at the surface of the sample. This experimental technique permits a successful decoupling of the sputtering and Cs introduction processes by avoiding the constraints imposed by an energetic Csþ ion bombardment. As a consequence, it becomes possible to optimize simultaneously the sensitivity of the analysis, by carefully adjusting the Cs concentration to its optimum value, and the depth resolution of the analysis, by choosing adequate primary bombardment conditions. In this paper, we will describe the new Cs0 column, which is based on an evaporation of pure metallic Cs, and outline its performances in terms of deposition rates, stability, beam dimensions and purity of the Cs deposit. # 2004 Elsevier B.V. All rights reserved. Keywords: Cesium evaporator; Cesium deposition; Cesium concentration
1. Introduction In order to enhance negative secondary ion emission, SIMS instruments are routinely equipped with a Csþ ion gun. As a matter of fact, the incorporation of Cs into the sample provokes a lowering of the work function and, as a consequence, an important increase of the negative ionization probability. On the other hand, the Csþ ion bombardment can be used to perform analyses in the MCsxþ mode, which is a well known technique allowing to get round the problems linked to the matrix effect. *
Corresponding author. Tel.: þ352-47-0261-514; fax: þ352-47-0261-549. E-mail address: [email protected] (T. Wirtz). 0169-4332/$ – see front matter # 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2004.03.180
A major drawback of using such a Csþ primary ion bombardment consists in the fact that this beam serves both for the incorporation of Cs in the material and for the sputtering of the surface. In this case, the primary bombardment conditions (mainly the angle and energy of impact) yield a distinct total sputtering yield and consequently determine the cesium surface concentration. It is thus impossible to separately choose the Cs concentration implanted in the sample, which is a crucial parameter determining the sensitivity of the analysis in the negative secondary ion mode as well as in the MCsxþ mode, and the energetic and angular parameters of the primary beam, which considerably affect major analytical characteristics such as the depth resolution.
T. Wirtz, H.-N. Migeon / Applied Surface Science 231–232 (2004) 940–944
In order to be able to perform analyses combining both an optimum sensitivity and an excellent depth resolution, an additional degree of freedom has to be obtained by separating completely the sputtering and Cs introduction phases during analyses. In this respect, we developed and installed on the Cation Mass Spectrometer (CMS), which is a prototype developed in our laboratory [1–4], a patented column that delivers an adjustable and collimated stream of neutral Cs atoms to be deposited on the surface of the sample while this one is being analyzed. Using this new column, it was possible to introduce an analysis technique consisting of a Xyþ ion bombardment (where X stands for any element excepting Cs) accompanied by a simultaneous deposition of Cs0 on the sample’s surface. This experimental technique permits a successful decoupling of the sputtering and Cs introduction processes by avoiding the constraints imposed by an energetic Csþ ion bombardment. When the optimum quantity of Cs is deposited in the form of neutral atoms on the surface of the sample, no sputtering process or atomic mixing of the target takes place and the depth resolution of the analysis depends solely on the characteristic conditions of the bombardment produced by the sputtering/analysis gun. In this paper, we will describe the new Cs0 column, which is based on an evaporation of pure metallic Cs, and outline its performances in terms of deposition rates, stability, beam dimensions and purity of the Cs deposit.
941
column delivering a stream of Cs from the evaporation of pure metallic Cs. 2.2. Description The Cs evaporator designed for the CMS machine comprises two separate parts (Fig. 1). Because of their respective positions with respect to the main chamber, the two parts will henceforth be referred to as the ‘‘internal’’ and ‘‘external’’ parts. The external part of the evaporator, which can be isolated from the main chamber of the CMS machine by means of a gate valve and be independently
2. The Cs0 evaporator 2.1. Operating principle To deposit alkaline metals, one almost exclusively uses getters that are heated by circulating an electric current in order to release the alkaline atoms. The use of such a getter for analytical purposes requiring a Cs ˚ /s, deposition rate of an order of magnitude of 1 A however, presents several major drawbacks consisting in a very limited lifetime, contamination of the analysis chamber due to a broad non-collimated beam and cleanliness of the Cs deposit due to traces of the chromates and reducing agents contained in the getter. In order to eliminate these major disadvantages of conventional getters, we developed as an alternative a
Fig. 1. Schematic overall view of the Cs0 evaporator. The external and internal parts are separated at the gate valve.
942
T. Wirtz, H.-N. Migeon / Applied Surface Science 231–232 (2004) 940–944
pumped or vented, consists of the actual evaporation block and a tube guiding the stream of gaseous Cs towards the internal part of the evaporator. The evaporation block is a solid cast stainless steel part containing a cylindrical housing in which the reservoir filled with the metallic Cs (several grams) slides. To obtain gaseous Cs, the housing of the reservoir (as well as the guiding tube) is wound round with a heating wire capable of supplying the thermal energy necessary for the evaporation of the Cs. The temperature of the reservoir is measured by means of a chromel–alumel thermocouple screwed to the end of the reservoir. The internal part of the evaporator serves to deliver the stream of neutral Cs onto the zone to be analyzed in the form of a jet of sufficiently reduced diameter to avoid any contamination of the analysis chamber. For this purpose, the end piece of the column ends in a cone with an escape hole 2 mm in diameter. All the pipes in the internal part can be heated by means of a further, independent heating wire. The temperature is monitored on the end piece of the gun by means of another thermocouple. The Cs0 column is mounted on the main chamber of the CMS machine by means of a union adjustable in distance and inclination to allow optimum positioning of the spot of Cs0 on the useful zone. The axis of the evaporator forms an angle of 458 with respect to the normal to the sample. 2.3. Characterization 2.3.1. Thermal behavior Because of their different weights and environments (vacuum for the internal part, atmosphere for the external part), the external and internal parts present fairly different thermal behaviors. It is thus necessary to apply a much greater power to the heating element of the external part than to that of the internal part when one wishes to raise the two tubes to the same temperature. The important mass of the evaporation block enables to produce a very stable heating temperature in this region, which is crucial in order to obtain a stable evaporation rate. On the other hand, the fact that the external part loses a considerable proportion of the heat received to the external environment via the evaporation block and via the bellows of the translation system,
the surface of which is comparable to that of a radiator, allows the temperature of the reservoir to be lowered quickly with a view to stopping the evaporator. 2.3.2. Cs0 deposition rates In order to be able to measure the stream of neutral Cs delivered by the evaporator, a mobile quartz microbalance system, which can be moved on a horizontal axis located at the same distance from the extraction optics as the sample during the analyses, was installed on the CMS machine. The sensor can thus be brought in front of the stream of Cs0 at the point where the sample is normally positioned. Fig. 2 represents the calibration curve enabling the stream of Cs0 to be adjusted by adapting the temperature of the reservoir. Independently from the power Pext applied, the internal part of the evaporator is raised during all the analyses to 110 8C in order to avoid any risk of condensation and obstruction in the various tubes. From this calibration curve it can be concluded that Cs0 deposition rates on the sample of ˚ /s can be reached with reasonthe desired order of 1 A able temperatures (75–85 8C) and corresponding heating powers (75–100 W). 2.3.3. Pressure conditions The reservoir temperature range required for the evaporator to output the necessary flow corresponds according to the Cs saturating pressure curve to a pressure range in the source of between 1 104 and 4 104 mbar [5,6]. This value appears to be realistic in view of the length (approximately 40 cm) and the small diameter (between 2 and 8 mm) of the pipe connecting the reservoir to the main chamber via which the pumping is carried out. In addition, we observe that the pressure in the analysis chamber remains at a quite acceptable level during the operation of the evaporator (4 108 to 1 107 mbar for deposition rates between 0.3 and ˚ /s). 3.5 A 2.3.4. Stability of deposition rates By recording the deposition rates indicated by the controller for various heating powers and therefore for various values of the evaporator flow rate, we determined a deposition rate stability DvD /vD ¼ 2% over 60 min.
T. Wirtz, H.-N. Migeon / Applied Surface Science 231–232 (2004) 940–944
943
3,5
Deposition rate (Å/s)
3,0 2,5 2,0 1,5 1,0 0,5 0,0 70
75
80
85
Temperature of reservoir (˚C) Fig. 2. Change in the Cs0 deposition rate on the sample as a function of the temperature of the reservoir.
2.3.5. Dimensions of Cs0 beam To evaluate the diameter of the spot of Cs0 on the sample, we measured the deposition rate with the quartz balance for various heating powers, while sweeping the diagonal of the beam with the sensor. Depending on the chosen criteria (Gaussian fit, width at mid-height), the diameter of the Cs0 spot can be evaluated at 7–9 mm. Given the dimensions of the sample holder (10 cm 7:5 cm), we can conclude that the whole stream of Cs0 delivered by the evaporator remains confined on this plate. 2.3.6. Purity of Cs0 deposit The cleanliness of the Cs deposit is a crucial point for the use of the Cs0 evaporator during the analyses. A contamination of the Cs vapor with impurities would lead to an increase in the detection limits for certain elements given that the signal of the element in question would be affected by a background noise of varying degrees of intensity. An analysis of the layer of Cs deposited on Si and AsGa samples was carried out by different means (SIMS using a Gaþ bombardment, SEM/EDX). The obtained spectra were compared with spectra of the same type produced beforehand on the untouched samples. This comparison shows that the spectra produced while the evaporator was depositing Cs on the surface are composed of the same peaks as the spectra of the untouched samples plus the typical
peaks characteristic for the presence of Cs atoms on the samples of Si or AsGa. However, there is no explicit trace of any contaminant.
3. Conclusion We have developed and installed on the Cation Mass Spectrometer (CMS) a column that delivers a collimated stream of neutral Cs atoms to be deposited on the surface of the sample to be analyzed by SIMS. The deposition rate of Cs0 can be adjusted by adapting the heating power applied to the evaporation block. Furthermore, no evidence for any contaminations contained in the Cs deposit could be found. Using this new column, it was possible to introduce an analysis technique consisting of a Xyþ ion bombardment (where X stands for any element excepting Cs) accompanied by a deposition of Cs0 at the surface of the sample. This experimental technique permits a successful decoupling of the sputtering and Cs introduction processes by avoiding the constraints imposed by an energetic Csþ ion bombardment. As a consequence, it becomes possible to optimize simultaneously the sensitivity of the analysis, by carefully adjusting the Cs concentration to its optimum value, and the depth resolution of the analysis, by choosing adequate primary bombardment conditions.
944
T. Wirtz, H.-N. Migeon / Applied Surface Science 231–232 (2004) 940–944
Acknowledgements This work was financially supported by the MCESR of Luxembourg.
References [1] T. Wirtz, H.-N. Migeon, H. Scherrer, Int. J. Mass Spectrom. 225 (2) (2003) 135.
[2] T. Wirtz, H.-N. Migeon, H. Scherrer, Appl. Surf. Sci. 203/204 (2003) 189. [3] T. Wirtz, B. Duez, H.-N. Migeon, H. Scherrer, Int. J. Mass Spectrom. 209 (2001) 57. [4] T. Mootz, B. Rasser, P. Sudraud, E. Niehuis, T. Wirtz, W. Bieck, H.-N. Migeon, in: A. Benninghoven, P. Bertrand, H.-N. Migeon, H.W. Werner (Eds.), Proceedings of the Secondary Ion Mass Spectrometry SIMS XII, Elsevier, Amsterdam, 2000, p. 233. [5] R.E. Honig, D.A. Kramer, RCA Rev. 30 (1969) 285. [6] C.B. Alcock, V.P. Itkin, M.K. Horrigan, Can. Metall. Q. 23 (1984) 23.
Development of a column delivering a collimated stream of Cs0 for SIMS purposes T. Wirtz*, H.-N. Migeon Laboratoire d’Analyse des Mate´riaux, Centre de Recherche Public—Gabriel Lippmann, 162A avenue de la Faı¨encerie, L-1511 Luxembourg, Luxembourg Available online 30 April 2004
Abstract In order to optimize SIMS analyses using the presence of reactive Cs for the purpose of enhancing negative secondary ion emission or working in the MCsxþ mode, we have developed a column that delivers a collimated and adjustable stream of neutral Cs atoms to be deposited on the surface of the sample while this one is being analyzed. Using this new column, it was possible to introduce an analysis technique consisting of a Xyþ ion bombardment (where X stands for any element excepting Cs) accompanied by a simultaneous deposition of Cs0 at the surface of the sample. This experimental technique permits a successful decoupling of the sputtering and Cs introduction processes by avoiding the constraints imposed by an energetic Csþ ion bombardment. As a consequence, it becomes possible to optimize simultaneously the sensitivity of the analysis, by carefully adjusting the Cs concentration to its optimum value, and the depth resolution of the analysis, by choosing adequate primary bombardment conditions. In this paper, we will describe the new Cs0 column, which is based on an evaporation of pure metallic Cs, and outline its performances in terms of deposition rates, stability, beam dimensions and purity of the Cs deposit. # 2004 Elsevier B.V. All rights reserved. Keywords: Cesium evaporator; Cesium deposition; Cesium concentration
1. Introduction In order to enhance negative secondary ion emission, SIMS instruments are routinely equipped with a Csþ ion gun. As a matter of fact, the incorporation of Cs into the sample provokes a lowering of the work function and, as a consequence, an important increase of the negative ionization probability. On the other hand, the Csþ ion bombardment can be used to perform analyses in the MCsxþ mode, which is a well known technique allowing to get round the problems linked to the matrix effect. *
Corresponding author. Tel.: þ352-47-0261-514; fax: þ352-47-0261-549. E-mail address: [email protected] (T. Wirtz). 0169-4332/$ – see front matter # 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.apsusc.2004.03.180
A major drawback of using such a Csþ primary ion bombardment consists in the fact that this beam serves both for the incorporation of Cs in the material and for the sputtering of the surface. In this case, the primary bombardment conditions (mainly the angle and energy of impact) yield a distinct total sputtering yield and consequently determine the cesium surface concentration. It is thus impossible to separately choose the Cs concentration implanted in the sample, which is a crucial parameter determining the sensitivity of the analysis in the negative secondary ion mode as well as in the MCsxþ mode, and the energetic and angular parameters of the primary beam, which considerably affect major analytical characteristics such as the depth resolution.
T. Wirtz, H.-N. Migeon / Applied Surface Science 231–232 (2004) 940–944
In order to be able to perform analyses combining both an optimum sensitivity and an excellent depth resolution, an additional degree of freedom has to be obtained by separating completely the sputtering and Cs introduction phases during analyses. In this respect, we developed and installed on the Cation Mass Spectrometer (CMS), which is a prototype developed in our laboratory [1–4], a patented column that delivers an adjustable and collimated stream of neutral Cs atoms to be deposited on the surface of the sample while this one is being analyzed. Using this new column, it was possible to introduce an analysis technique consisting of a Xyþ ion bombardment (where X stands for any element excepting Cs) accompanied by a simultaneous deposition of Cs0 on the sample’s surface. This experimental technique permits a successful decoupling of the sputtering and Cs introduction processes by avoiding the constraints imposed by an energetic Csþ ion bombardment. When the optimum quantity of Cs is deposited in the form of neutral atoms on the surface of the sample, no sputtering process or atomic mixing of the target takes place and the depth resolution of the analysis depends solely on the characteristic conditions of the bombardment produced by the sputtering/analysis gun. In this paper, we will describe the new Cs0 column, which is based on an evaporation of pure metallic Cs, and outline its performances in terms of deposition rates, stability, beam dimensions and purity of the Cs deposit.
941
column delivering a stream of Cs from the evaporation of pure metallic Cs. 2.2. Description The Cs evaporator designed for the CMS machine comprises two separate parts (Fig. 1). Because of their respective positions with respect to the main chamber, the two parts will henceforth be referred to as the ‘‘internal’’ and ‘‘external’’ parts. The external part of the evaporator, which can be isolated from the main chamber of the CMS machine by means of a gate valve and be independently
2. The Cs0 evaporator 2.1. Operating principle To deposit alkaline metals, one almost exclusively uses getters that are heated by circulating an electric current in order to release the alkaline atoms. The use of such a getter for analytical purposes requiring a Cs ˚ /s, deposition rate of an order of magnitude of 1 A however, presents several major drawbacks consisting in a very limited lifetime, contamination of the analysis chamber due to a broad non-collimated beam and cleanliness of the Cs deposit due to traces of the chromates and reducing agents contained in the getter. In order to eliminate these major disadvantages of conventional getters, we developed as an alternative a
Fig. 1. Schematic overall view of the Cs0 evaporator. The external and internal parts are separated at the gate valve.
942
T. Wirtz, H.-N. Migeon / Applied Surface Science 231–232 (2004) 940–944
pumped or vented, consists of the actual evaporation block and a tube guiding the stream of gaseous Cs towards the internal part of the evaporator. The evaporation block is a solid cast stainless steel part containing a cylindrical housing in which the reservoir filled with the metallic Cs (several grams) slides. To obtain gaseous Cs, the housing of the reservoir (as well as the guiding tube) is wound round with a heating wire capable of supplying the thermal energy necessary for the evaporation of the Cs. The temperature of the reservoir is measured by means of a chromel–alumel thermocouple screwed to the end of the reservoir. The internal part of the evaporator serves to deliver the stream of neutral Cs onto the zone to be analyzed in the form of a jet of sufficiently reduced diameter to avoid any contamination of the analysis chamber. For this purpose, the end piece of the column ends in a cone with an escape hole 2 mm in diameter. All the pipes in the internal part can be heated by means of a further, independent heating wire. The temperature is monitored on the end piece of the gun by means of another thermocouple. The Cs0 column is mounted on the main chamber of the CMS machine by means of a union adjustable in distance and inclination to allow optimum positioning of the spot of Cs0 on the useful zone. The axis of the evaporator forms an angle of 458 with respect to the normal to the sample. 2.3. Characterization 2.3.1. Thermal behavior Because of their different weights and environments (vacuum for the internal part, atmosphere for the external part), the external and internal parts present fairly different thermal behaviors. It is thus necessary to apply a much greater power to the heating element of the external part than to that of the internal part when one wishes to raise the two tubes to the same temperature. The important mass of the evaporation block enables to produce a very stable heating temperature in this region, which is crucial in order to obtain a stable evaporation rate. On the other hand, the fact that the external part loses a considerable proportion of the heat received to the external environment via the evaporation block and via the bellows of the translation system,
the surface of which is comparable to that of a radiator, allows the temperature of the reservoir to be lowered quickly with a view to stopping the evaporator. 2.3.2. Cs0 deposition rates In order to be able to measure the stream of neutral Cs delivered by the evaporator, a mobile quartz microbalance system, which can be moved on a horizontal axis located at the same distance from the extraction optics as the sample during the analyses, was installed on the CMS machine. The sensor can thus be brought in front of the stream of Cs0 at the point where the sample is normally positioned. Fig. 2 represents the calibration curve enabling the stream of Cs0 to be adjusted by adapting the temperature of the reservoir. Independently from the power Pext applied, the internal part of the evaporator is raised during all the analyses to 110 8C in order to avoid any risk of condensation and obstruction in the various tubes. From this calibration curve it can be concluded that Cs0 deposition rates on the sample of ˚ /s can be reached with reasonthe desired order of 1 A able temperatures (75–85 8C) and corresponding heating powers (75–100 W). 2.3.3. Pressure conditions The reservoir temperature range required for the evaporator to output the necessary flow corresponds according to the Cs saturating pressure curve to a pressure range in the source of between 1 104 and 4 104 mbar [5,6]. This value appears to be realistic in view of the length (approximately 40 cm) and the small diameter (between 2 and 8 mm) of the pipe connecting the reservoir to the main chamber via which the pumping is carried out. In addition, we observe that the pressure in the analysis chamber remains at a quite acceptable level during the operation of the evaporator (4 108 to 1 107 mbar for deposition rates between 0.3 and ˚ /s). 3.5 A 2.3.4. Stability of deposition rates By recording the deposition rates indicated by the controller for various heating powers and therefore for various values of the evaporator flow rate, we determined a deposition rate stability DvD /vD ¼ 2% over 60 min.
T. Wirtz, H.-N. Migeon / Applied Surface Science 231–232 (2004) 940–944
943
3,5
Deposition rate (Å/s)
3,0 2,5 2,0 1,5 1,0 0,5 0,0 70
75
80
85
Temperature of reservoir (˚C) Fig. 2. Change in the Cs0 deposition rate on the sample as a function of the temperature of the reservoir.
2.3.5. Dimensions of Cs0 beam To evaluate the diameter of the spot of Cs0 on the sample, we measured the deposition rate with the quartz balance for various heating powers, while sweeping the diagonal of the beam with the sensor. Depending on the chosen criteria (Gaussian fit, width at mid-height), the diameter of the Cs0 spot can be evaluated at 7–9 mm. Given the dimensions of the sample holder (10 cm 7:5 cm), we can conclude that the whole stream of Cs0 delivered by the evaporator remains confined on this plate. 2.3.6. Purity of Cs0 deposit The cleanliness of the Cs deposit is a crucial point for the use of the Cs0 evaporator during the analyses. A contamination of the Cs vapor with impurities would lead to an increase in the detection limits for certain elements given that the signal of the element in question would be affected by a background noise of varying degrees of intensity. An analysis of the layer of Cs deposited on Si and AsGa samples was carried out by different means (SIMS using a Gaþ bombardment, SEM/EDX). The obtained spectra were compared with spectra of the same type produced beforehand on the untouched samples. This comparison shows that the spectra produced while the evaporator was depositing Cs on the surface are composed of the same peaks as the spectra of the untouched samples plus the typical
peaks characteristic for the presence of Cs atoms on the samples of Si or AsGa. However, there is no explicit trace of any contaminant.
3. Conclusion We have developed and installed on the Cation Mass Spectrometer (CMS) a column that delivers a collimated stream of neutral Cs atoms to be deposited on the surface of the sample to be analyzed by SIMS. The deposition rate of Cs0 can be adjusted by adapting the heating power applied to the evaporation block. Furthermore, no evidence for any contaminations contained in the Cs deposit could be found. Using this new column, it was possible to introduce an analysis technique consisting of a Xyþ ion bombardment (where X stands for any element excepting Cs) accompanied by a deposition of Cs0 at the surface of the sample. This experimental technique permits a successful decoupling of the sputtering and Cs introduction processes by avoiding the constraints imposed by an energetic Csþ ion bombardment. As a consequence, it becomes possible to optimize simultaneously the sensitivity of the analysis, by carefully adjusting the Cs concentration to its optimum value, and the depth resolution of the analysis, by choosing adequate primary bombardment conditions.
944
T. Wirtz, H.-N. Migeon / Applied Surface Science 231–232 (2004) 940–944
Acknowledgements This work was financially supported by the MCESR of Luxembourg.
References [1] T. Wirtz, H.-N. Migeon, H. Scherrer, Int. J. Mass Spectrom. 225 (2) (2003) 135.
[2] T. Wirtz, H.-N. Migeon, H. Scherrer, Appl. Surf. Sci. 203/204 (2003) 189. [3] T. Wirtz, B. Duez, H.-N. Migeon, H. Scherrer, Int. J. Mass Spectrom. 209 (2001) 57. [4] T. Mootz, B. Rasser, P. Sudraud, E. Niehuis, T. Wirtz, W. Bieck, H.-N. Migeon, in: A. Benninghoven, P. Bertrand, H.-N. Migeon, H.W. Werner (Eds.), Proceedings of the Secondary Ion Mass Spectrometry SIMS XII, Elsevier, Amsterdam, 2000, p. 233. [5] R.E. Honig, D.A. Kramer, RCA Rev. 30 (1969) 285. [6] C.B. Alcock, V.P. Itkin, M.K. Horrigan, Can. Metall. Q. 23 (1984) 23.