Surface Science 528 (2003) 261–265 www.elsevier.com/locate/susc
Construction and evaluation of an electron-ion coincidence apparatus using a large transmission coaxially symmetric mirror electron energy analyzer Kouji Isari a
a,*
, Eiichi Kobayashi b, Kazuhiko Mase b, Kenichiro Tanaka
a
Department of Physical Science, Hiroshima University, 1-3-1 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8526, Japan b Institute of Materials Structure Science, 1-1 Oho, Tsukuba 305-0801, Japan
Abstract We have developed an electron ion coincidence (EICO) apparatus using a coaxially symmetric mirror electron energy analyzer with a large transmission. The electron energy analyzer originally proposed by Kai Siegbahn in 1997 had a weak point that the performance is degraded by disturbance of the electric field near the end plates. We have improved this by adopting compensation electrodes. Simulation with SIMION 3D version 7.0 predicts that the energy resolution (FWHM) is E=DE ¼ 300 for a pointed beam and a solid angle of 1.2 sr. This value is about three times better than the predicted value for the cylindrical mirror analyzer (CMA) used in the previous EICO apparatus. The measured electron energy resolution is E=DE ¼ 120 for a beam size of 2 mm 1 mm, which is about 1.5 times better than that for the previous CMA. Auger electron photoion coincidence measurement of condensed H2 O at 4a1 O 1s resonance showed that the measurement time required for a certain coincidence ion counts is reduced by one order of magnitude in comparison with the previous EICO. Ó 2002 Published by Elsevier Science B.V. Keywords: Photon stimulated desorption (PSD); Photoelectron spectroscopy; Auger electron spectroscopy; Synchrotron radiation photoelectron spectroscopy
1. Introduction When a surface is irradiated by electrons or X-rays, ions are known to be desorbed by the following three-step processes, that is, (1) a coreelectron transition leaving a core hole (0.1 fs), (2) an Auger transition leaving multiple valence holes (1–10 fs), and (3) decay of the multi-hole state, causing ion desorption (10–100 fs) (Auger stimulated ion desorption (ASID) mechanism) [1,2]. To
*
Corresponding author.
clarify details of ASID mechanism measurements of ion desorption yield for the selected core-excitation-final-states or the selected Auger-final-states are required. For this reason we have developed electron ion coincidence (EICO) apparatus [3]. The EICO apparatus is composed of an electron energy analyzer and a time-of-flight ion mass spectrometer (TOF-MS). A sample surface is excited by synchrotron radiation, and energy of the emitted electrons is detected with the electron energy analyzer, while the desorbed ions were accelerated towards the TOF-MS. The ion counts were recorded, as a function of TOF difference between
0039-6028/02/$ - see front matter Ó 2002 Published by Elsevier Science B.V. doi:10.1016/S0039-6028(02)02642-0
262
K. Isari et al. / Surface Science 528 (2003) 261–265
the energy-selected electron and the ion with a multichannel scaler (MCS) by taking the electron signal as the starting trigger. The ion desorbed in coincidence with the detected electron gives a coincidence signal at a TOF specific for the mass number, while the ion irrelevant to the electron increases the background level. With EICO apparatus we have investigated the details of ion desorption mechanism of various samples such as condensed H2 O, condensed CH3 CN, poly-methylmethacrylate thin film, H2 O/ Si(1 0 0), CaF2 (1 1 1) [4–6]. EICO apparatus is useful also for excitation site-specific ion desorption investigations of molecules containing multiple sites with a different chemical environment [4,5,7], and study of core-level shift of the site responsible for ion desorption [5,8]. We have developed three types of EICO analyzer using a cylindrical mirror analyzer (CMA) so far [3–5]. Since the coincidence signal counts are linearly dependent on the collection efficiency of electrons and ions [9], the solid angles of CMA and TOF-MS were improved at every remodeling. The performance of the past EICO apparatus, however, was limited by the CMA (solid angle 6 1:1 sr and E=DE 6 100) [5]. In 1997 Siegbahn et al. have developed a coaxially symmetric mirror electron energy analyzer, the solid angle and the resolution of which are higher than those of CMA [10]. In this article we describe details and performance of a new EICO apparatus using a modified coaxially symmetric mirror electron energy analyzer.
2. Modified coaxially symmetric mirror electron energy analyzer To develop the coaxially symmetric mirror electron energy analyzer, Siegbahn et al. considered the Laplace equation (1) 1 o ou o2 u r þ 2 ¼0 r or or oz
ð1Þ
which describes potential (u) of the coaxially symmetric electric field, where r is the distance from the axis and z is the coordinate in the direction of the axis [10]. Then the coaxially symmetric potential is given by Eq. (2).
u ¼ a ln r þ b
r2 z2 2
þ cz þ d
ð2Þ
CMA, which is a most popular coaxially symmetric mirror analyzer, corresponds to the case of b ¼ c ¼ 0, u ¼ a ln r þ d
ð3Þ
and the inner and the outer electrodes take the form of a cylinder. On the other hand, Siegbahn et al. found that the convergence of the electron trajectory is better in the electric field given by b ¼ a and c ¼ 0 2 r 2 u ¼ a ln r a ð4Þ z þd 2 than that of a CMA, and developed a coaxially symmetric mirror analyzer which consists of the inner and the outer electrodes [10]. The analyzer, however, had a weak point that the performance is degraded by the disturbance of the electric field near the end plates. We have improved this and developed a modified analyzer, which consists of an inner electrode, an outer electrode, three sets of compensation electrodes, a slit with a diameter of 0.8 mm, micro channel plates (MCP), and a magnetic shield [11]. 3. Electron-ion coincidence analyzer Figs. 1 and 2 show the cross section and a photograph of the developed EICO analyzer using a modified coaxially symmetric mirror electron energy analyzer, respectively. The EICO analyzer is set up on a conflat flange of a 203-mm-diameter, which installs a positioning mechanism. The distance between the sample and the front end of the analyzer is 2.4 mm. The incidence angle of synchrotron radiation is 84° from the surface normal. The glancing incidence is advantageous for desorption induced by p-polarized radiation [12]. We developed a compact TOF-MS with a diameter of 26.0 mm and installed it coaxially in the inside of the electron energy analyzer (see, Fig. 1). The TOFMS is composed of an electric field shield, an ionextraction electrode, a drift tube, and MCP. The transmittance of the three meshes inserted perpendicular to the axis of the TOF tube is 0.47, and the ion detection efficiency of the MCP is 0.60. The
K. Isari et al. / Surface Science 528 (2003) 261–265
Fig. 1. Cross section based on the draft of the developed EICO analyzer using a coaxially symmetric mirror electron energy analyzer and a compact TOF-MS [11]. There are four windows on the inner electrode through which the electrons fly. These are covered by meshes with a transmittance of 71% to avoid distortion of the electric field. The total azimuthal angle of the windows is 307°. The isoelectric lines simulated by the SIMION 3D version 7.0 are shown for the inner electrode and the magnetic shield setting at 0 V, the outer electrode at )100 V, the ion-extraction electrode at )30 V, and the drift tube at )300 V. Electron trajectories are also shown for the pointed beam and the electron kinetic energy of 181.7 eV and the emission angles of 49.1°–66.1° from the surface normal (the solid angle is 1.2 sr). Ion trajectories are also shown for the kinetic energy of 5 eV and the emission angles of 0°–25° from the surface normal. The ion acceptance angle can be increased voltage at the ionextraction electrode.
Fig. 2. Photograph of the coaxially symmetric mirror electron energy analyzer. A TOF-MS is built coaxially in the inside of analyzer.
263
distance between the sample and the front end of the TOF-MS is 3.4 mm, which is much shorter than the value of 6.0 mm of the previous EICO analyzer using a CMA [5]. Fig. 3 shows the electron trajectories at the slit of the coaxially symmetric mirror analyzer of the new EICO apparatus for a pointed beam and an ion-extraction voltage of )30 V. The distance between the sample and the slit of the coaxially symmetric mirror analyzer is 143.4 mm. For the comparison, the electron trajectories are also shown for the previous EICO analyzer using a CMA for a pointed beam and an ion-extraction voltage of )100 V. The electron acceptance angle from the surface normal is 49.1°–66.1°, which is larger than that of the CMA (37°–49°). So the new EICO analyzer is more surface-sensitive than the previous one. The designed resolution and the
Fig. 3. Electron trajectories around the slit of (a) the CMA of the previous EICO apparatus and (b) the coaxially symmetric mirror analyzer of the new one. The cross sections are based on the drafts. The electron kinetic energies are 181.1, 181.7 and 182.3 eV. The diameter of the slit and the width of the space in front of the slit are 10.0 and 1.0 mm for the CMA, and 0.8 and 0.5 mm for the coaxially symmetric mirror analyzer, respectively. The ion-extraction electrode is set at )100 V for the previous EICO and )30 V for the new one.
K. Isari et al. / Surface Science 528 (2003) 261–265
solid angle are estimated as E=DE ¼ 300 and 1.2 sr, respectively. The convergence of the electron trajectories at the slit are not degraded with the ion-extraction voltage owing to the short distance between the sample and the TOF-MS, and to the large acceptance angle of the coaxially symmetric mirror analyzer. This result shows that the modified coaxially symmetric mirror analyzer is more suitable for EICO apparatus than a CMA. Without the compensation electrodes, the electric field near the end plates is seriously disturbed. In this case the simulation showed that the electron energy resolution and the collection solid angle are degraded to E=DE ¼ 50 and 0.3 sr, respectively.
4. Evaluation of the performance
Photoelectron counts (cps)
The performance of the new EICO analyzer was evaluated at the BL-8A of the synchrotron radiation facility in the Institute of Materials Structure Science (PF). Fig. 4 shows a photoelectron spectrum of a clean Si(1 1 1) surface at ht ¼ 269 eV. The photon energy resolution E=DE is better than 1000. The intensity of synchrotron radiation was monitored with a gold mesh with a transmittance of 78%. The surface was cleaned by direct current heating at 1100 °C for 1 s under ultrahigh vacuum (1 109 Torr during the heating). The surface showed a clear 7 7 LEED pattern. The spot size of the synchrotron radiation was about 2 mm 1 mm on the surface. Based on the curve fitting of the
5000
Clean Si(111)
4000
Si 2p3/2
3000 2000
Si 2p1/2
FW HM 1.35eV
1000 0
Si:2p peak by a Voigt function we estimated the actual resolution as E=DE ¼ 120. This value is poorer than the predicted one for the pointed beam because the spot size of the synchrotron radiation is large, and because the error of the processing and assembly of the electrodes are large [13]. However, the value is still better by a factor of about 1.5 than that for the previous EICO analyzer using a CMA. Fig. 5 shows an Auger electron photoion coincidence (AEPICO) spectrum of condensed H2 O at 100 K measured at 4a1 O 1s resonance (ht ¼ 533:6 eV). Fig. 6 shows a Hþ AEPICO yield spectrum of condensed H2 O at 4a1 O 1s. Improvement of the resolution usually degrades the signal-to-background (S/B) ratio in the EICO spectroscopy, because the number of detected electron is reduced for a higher resolution. Thus the coincidence ion signal is reduced, while the ion desorbed in coincidence with the undetected electron increases the background. The S/B ratio in Figs. 5 and 6, however, is much better than that obtained with the previous EICO apparatus [5]. In order to evaluate the performance of the EICO apparatus, we introduced the following standard independent of the photon intensity, that is ðCoincidence ion counts ½cps Þ =ðphotoion counts ½cps Þ 200 +
Ion signal (counts)
264
H 296 ns
150
100
50
0
0
500
1000
1500
2000
Electron-ion TOF difference (ns) 160
162 164 166 168 Electron kinetic energy (eV)
Fig. 4. Photoelectron spectrum of a clean Si(1 1 1) surface at ht ¼ 269 eV.
Fig. 5. Typical AEPICO spectrum of condensed H2 O at 4a1 O 1s resonance (ht ¼ 533:6 eV) for the electron kinetic energy of 506 eV measured with the present EICO apparatus. Time of measurement was 600 s. The count rates of the Hþ coincidence, photoelectron and photoion count rates are 0:772 0:035, 588 20 and 24 5 cps, respectively.
600
Auger electron spectrum
400
400 200 200
AEPICO yield spectrum 0 0
460
480
500
520
0
+ H AEPICO yield (counts)
Auger electron counts (cps)
K. Isari et al. / Surface Science 528 (2003) 261–265
Electron kinetic energy (eV) Fig. 6. Auger and Hþ AEPICO yield spectra of condensed H2 O at 4a1 O 1s resonance at ht ¼ 533:6 eV. The mesh current monitoring the photon intensity was 23:37 0:05 pA.
This value reflects coincidence ion detection efficiency, which depends on the collection efficiency of the electron energy analyzer and the TOF-MS as well as photon energy resolution and photon incidence angle. The value obtained from Fig. 5 was 0.032 at ht ¼ 533:6 eV for the electron kinetic energy of 506 eV, which is one order of magnitude better than that for the previous EICO apparatus for photon energy resolution of E=DE ¼ 500 and photon incidence angle of 30° from the surface normal [5], that is, the measurement time required for a certain coincidence ion counts is reduced by one order of magnitude. We have applied the EICO apparatus for excitation site-specific ion desorption study of Si(1 1 1) surfaces fluorinated by XeF2 [14]. Owing to the improved electron energy resolution and the reduced measurement time, we have succeeded to investigate the details of Fþ desorption induced by Si:2p ionization of the SiF, SiF2 and SiF3 sites as a function of the XeF2 exposure.
Acknowledgements We express our gratitude for Mr. K. Kudara, Mr. K. Kusaba, Mr. Y. Matsumoto, Mr. K.
265
Morita, (Hiroshima Univ.) and Mr. T. Ibe (Chiba Univ.) for their valuable supports. This work was supported by Research for the Future Program (Photoscience; Molecular Knife––Control of Chemical Reactions by Core Excitation, JSPSRFTF98P01202) of Japan Society for the Promotion of Science, and by Grant-in-Aids for Scientific Research on the Priority Area ‘‘Manipulation of Atoms and Molecules by Electronic Excitation’’ (11222206) and (12555007) from the Japanese Ministry of Education, Culture, Sports, Science and Technology.
References [1] M.L. Knotek, P.J. Feibelman, Phys. Rev. Lett. 40 (1978) 964. [2] R. Franchy, D. Menzel, Phys. Rev. Lett. 43 (1979) 865. [3] K. Mase, M. Nagasono, S. Tanaka, M. Kamada, T. Urisu, Y. Murata, Rev. Sci. Inst. 68 (1997) 1703. [4] K. Mase, S. Tanaka, Jpn. J. Appl. Phys. Suppl. 38-1 (1999) 233. [5] K. Mase, S. Tanaka, S. Nagaoka, T. Urisu, Surf. Sci. 451 (2000) 143. [6] K. Tanaka, E.O. Sako, E. Ikenaga, K. Isari, S.A. Sardar, S. Wada, T. Sekitani, K. Mase, N. Ueno, J. Electron. Spectros. Rel. Phenom. 119 (2001) 255. [7] S. Nagaoka, K. Mase, M. Nagasono, S. Tanaka, T. Urisu, J. Ohshita, J. Chem. Phys. 107 (1997) 10751. [8] S. Tanaka, K. Mase, M. Nagasono, S. Nagaoka, M. Kamada, E. Ikenaga, T. Sekitani, S. Tanaka, Jpn. J. Appl. Phys. 39 (2000) 4489. [9] M.L. Knotek, J.W. Rabalais, in: W. Brenig, D. Menzel, (Eds.), Desorption Induced by Electronic Transitions, DIET-II, Springer Series in Surface Sciences, vol. 4, Springer, Berlin, 1985, p. 77. [10] K. Siegbahn, N. Kholine, G. Golikov, Nucl. Instr. and Meth. in Phys. Res. A 384 (1997) 563. [11] K. Isari, K. Tanaka, S. Nagaoka, T. Gejo, H. Yoshida, K. Mase, J. Vac. Soc. Jpn. (in Japanese), in preparation. [12] D. Coulman, A. Puschmann, U. H€ ofer, H.-P. Steinr€ uck, W. Wurth, P. Feulner, D. Menzel, J. Chem. Phys. 93 (1990) 58. [13] D. Briggs, M.P. Seah (Eds.), Practical surface analysis by Auger and X-ray Photoelectron Spectroscopy, John Wiley & Sons Ltd, 1983. [14] E. Kobayashi, K. Isari, K. Mase, Surf. Sci. 528 (2003) 255.