1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %% ws-procs9x6.tex : 20-9-2004
3: %% Text file for Proceedings Trim Size [9in x 6in] written in Latex2E.
4: %% The content, structure, format and layout of this style file is the
5: %% property of World Scientific Publishing Co. Pte. Ltd.
6: %% Copyright 1995, 2002 by World Scientific Publishing Co.
7: %% All rights are reserved.
8: %%
9: %% Proceedings Trim Size: 9in x 6in
10: %% Text Area: 7.35in (include runningheads) x 4.5in
11: %% Main Text is 10/13pt
12: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
13:
14: %% Use \tbl{...} command for table caption i.e. to fit table width.
15: %% Use \caption{...} command for figure caption.
16: %\documentclass[draft]{ws-procs9x6}
17: \documentclass{ws-procs9x6}
18:
19: \begin{document}
20:
21: \title{Atomic Hydrogen Cleaning of Polarized G\lowercase{a}A\lowercase{s} Photocathodes\footnote{\uppercase{T}his work was supported in part by the \uppercase{U.~S.~D}epartment of \uppercase{E}nergy under contract numbers \uppercase{DE-AC}02-76\uppercase{SF}00515 (\uppercase{SLAC}), and \uppercase{DE-AC}02-76\uppercase{ER}00881 (\uppercase{UW}).}}
22:
23: \author{T. Maruyama$^\dagger$, D.-A. Luh$^\dagger$, A. Brachmann$^\dagger$,
24: J.~E. Clendenin$^\dagger$, E.~L. Garwin$^\dagger$,
25: S. Harvey$^\dagger$, R.~E. Kirby$^\dagger$,
26: C.~Y. Prescott$^\dagger$, \\
27: and R. Prepost$^\ddagger$}
28: \address{
29: $^\dagger$Stanford Linear Accelerator Center, Menlo Park, California~94025 \\
30: $^\ddagger$Department of Physics, University of Wisconsin, Madison, Wisconsin~53706}
31:
32: \maketitle
33:
34: \abstracts{
35: Atomic hydrogen cleaning followed by heat cleaning at 450$^\circ$C
36: was used to prepare negative-electron-affinity GaAs photocathodes.
37: When hydrogen ions were eliminated, quantum efficiencies of 15\% were obtained
38: for bulk GaAs cathodes, higher than the results obtained using conventional
39: 600$^\circ$C heat cleaning.
40: The low-temperature cleaning technique was successfully applied to thin,
41: strained GaAs cathodes used for producing highly polarized electrons.
42: No depolarization was observed even when the optimum cleaning time of
43: about 30 seconds was extended by a factor of 100.
44: }
45:
46: \section{Introduction}
47: Recently the high-gradient-doping technique has been applied to
48: photocathode structures to successfully overcome the surface charge limit
49: effect while maintaining high spin polarization.
50: The high-gradient-doping technique consists of a thin (10 nm),
51: very-highly-doped (5$\times$10$^{19}$ cm$^{-3}$) surface layer with
52: a lower density doping (5$\times$10$^{17}$ cm$^{-3}$) in the remaining
53: active layer.
54: High dopant density promotes recombination of the minority carriers trapped
55: at the surface. Trapped carriers increase the surface barrier in proportion
56: to the arrival
57: rate of photoexcited conduction band (CB) electrons. Because CB electrons
58: depolarize as they diffuse to the surface of heavily doped materials,
59: the highly doped layer must be very thin, typically no more than a few
60: nanometers.
61: However, to achieve high quantum efficiencies, an negative-alectron-affinity (NEA) surface is required,
62: which in turn must be prepared on an atomically clean surface.
63: The conventional way to achieve a surface free of all surface oxides and
64: carbon-related contaminants is to heat the crystal to 600$^\circ$C for
65: about 1 hour. After only about 2 hours at this temperature, diffusion of
66: the dopant in the thin,
67: highly-doped layer results in the re-appearance of the surface charge limit
68: effect.
69: Therefore, high temperature heat cleaning should be avoided.
70:
71: Atomic hydrogen cleaning (AHC) is a well known technique for removing oxides
72: and carbon-related contaminants at relatively low temperatures.
73: While As-oxides and Ga$_2$O-like oxides are liberated at temperatures less
74: than 450$^\circ$C, the removal of Ga$_2$O$_3$-like oxides requires a higher
75: temperature. Under atomic hydrogen irradiation, Ga$_2$O$_3$-like oxides are
76: converted to more volatile Ga$_2$O-like oxides.
77: On the other hand, it has been well demonstrated that atomic hydrogen can
78: passivate both shallow donor and acceptor impurities.
79: The passivation rate increases rapidly with the doping concentration.
80: Since the band-bending in the photocathode is controlled by the $p-$type
81: doping, the dopant passivation may have an adverse effect on QE.
82:
83: In the present paper, a systematic study of AHC
84: in a vacuum-loading system is reported. The AHC system and the
85: associated analysis system
86: remain under UHV, while the sample is introduced in the UHV system
87: through a loading chamber, and transferred between the AHC and analysis
88: systems under vacuum \cite{tm}.
89:
90: \section{Experiment}
91: Two types of GaAs samples were used. Samples (13 $\times$ 13 mm$^2$) cut
92: from Zn-doped (1$\times$10$^{19}$ cm$^{-3}$) bulk GaAs (001) wafers were used
93: for optimizing the AHC conditions. Strained GaAs samples
94: with the active 100-nm GaAs layer Zn-doped at 5$\times$10$^{18}$ cm$^{-3}$
95: were used for studying the AHC effect on polarization.
96: Prior to installation
97: in the loading chamber, a sample was degreased in a boiling solution of
98: trichloroethylene and
99: chemically cleaned in NH$_4$OH to remove native oxides on the surface.
100: Since NH$_4$OH etches only oxides without disturbing the stoichiometry of GaAs,
101: it was used for the epitaxial photocathodes as well.
102: Some samples were installed without the NH$_4$OH cleaning
103: to intentionally leave native oxides on the surface.
104:
105: The experiments were carried out in a three-chamber UHV
106: system consisting of a loading chamber, an AHC system and an analysis system
107: called Cathode Test System (CTS).
108: The AHC system was equipped with an rf plasma discharge source, a heater tower,
109: and a linear translator. The surface temperature
110: during AHC was maintained at 300 - 350$^\circ$C.
111: The heater tower was electrically isolated from the AHC system so that
112: a bias voltage could be applied to the GaAs sample during AHC.
113: Atomic hydrogen was produced by dissociating molecular hydrogen in
114: a 2.5 cm diameter Pyrex glassware surrounded by a helical rf resonator
115: following the design used at Jefferson Lab. \cite{sinclair}
116: To study the effect of hydrogen ions generated by the rf dissociator,
117: the GaAs sample could be
118: biased negatively to enhance ion collection. To reduce
119: the ion current, a transverse magnetic field was applied at the exit of the
120: dissociator using a permanent magnet.
121: With the magnet in place, the ion current was negligible ( $<$ 1 nA).
122:
123: Activation to an NEA surface, and measurement of
124: QE and polarization were made in the CTS.
125: After AHC, the sample was transferred to the CTS under vacuum
126: when the AHC system pressure dropped
127: to a few 10$^{-8}$ Torr, typically within 30 minutes after AHC.
128: The cathode activation method used to obtain an NEA surface
129: consisted of heat cleaning
130: for 1 hour, cool-down for an hour, followed by
131: application of cesium until the photo-yield peaked, and then cesium and
132: nitrogen-trifluoride co-deposition until the photo-yield was again maximized.
133:
134: AHC was performed on NH$_4$OH-cleaned bulk GaAs samples under three
135: different conditions that control the ion current: 1)
136: with no bias and no magnet, 2) with a negative 88-Volt bias and no magnet to enhance
137: the ion effect, and 3) with magnet and no bias to eliminate the ion effect.
138: For each AHC, a fresh cathode sample was used and only the AHC time was
139: varied while all
140: other conditions were fixed. After AHC, the sample was transferred to the CTS
141: under vacuum,
142: heat-cleaned at 450$^\circ$C, activated, and the QE measured.
143: Figure 1 shows the QE at 670 nm as a function of the AHC time.
144: QEs as high as 16\% were obtained with only 15 - 30 seconds of AHC.
145: The QE decreased with prolonged cleaning, yielding only 1.8\% after 40 minutes
146: of AHC. The QE was higher when the ion current was eliminated
147: using the magnet, and lower when the ion current was enhanced using the
148: negative bias.
149: The data indicate that the excessive absorption of atomic hydrogen in GaAs
150: is detrimental for the QE.
151: Under atomic hydrogen irradiation, the native oxides on the surface are
152: converted to more volatile oxides and get liberated.
153: If the irradiation continues, atomic hydrogen is absorbed in the GaAs,
154: passivating the $p$-type dopants in the band bending region.
155: Since the doping concentration at the surface controls
156: the band bending, dopant passivation raises the vacuum level,
157: resulting in a lower QE.
158: Acceptor passivation by hydrogen proceeds through ion pair formation of
159: negatively charged acceptors and positively charged hydrogen ions.
160: \cite{pearton}
161: Ions are more effective than atomic hydrogen at passivating the dopants.
162:
163: AHC was also performed on bulk GaAs samples without the NH$_4$OH cleaning.
164: QEs as high as 14\% were obtained after 1 hour of AHC
165: (open square in Figure 1).
166: The sample with native oxides produced a high QE after prolonged AHC,
167: indicating the oxide layer was protecting the GaAs surface from impinging
168: atomic hydrogen. \cite{ide}
169: This indicates the AHC time must be optimized depending on the
170: oxide level on the GaAs surface.
171: The NH$_4$OH etching establishes a reproducible level of native oxides.
172:
173: \begin{figure}[ht]
174: \centerline{\epsfxsize=2.5in\epsfbox{ahc_fig1.eps}}
175: \caption{QE at 670 nm as a function of the AHC time. Three different
176: conditions are used to control ions: solid circles are no bias and with
177: magnet, squares are no bias and without magnet, and triangles are
178: with -88 V negative bias without magnet. One sample (open square) was
179: not cleaned in NH$_4$OH.}
180: \end{figure}
181:
182: Studies at Jefferson Lab \cite{baylac} indicate that a significant
183: depolarization may occur as a result of long exposures to atomic deuterium.
184: The electron polarization was measured in the present experiment
185: as a function of the AHC time using thin strained photocathodes.
186: The sample was heat-cleaned at 570$^\circ$C without AHC and the polarization
187: was measured. Then, a sequence consisting of 15 minutes of AHC followed by
188: polarization measurements was repeated four times for the same sample.
189: The AHC was performed with the magnet and no bias voltage.
190: Figure 2 shows the polarization spectrum
191: for 1) no AHC, 2) 30 minutes AHC, and 3) 60 minutes AHC. All three data
192: sets are consistent within the statistical errors. No depolarization
193: was observed after 60 minutes, which is $\sim$100 times longer exposure
194: time than the optimum time, of AHC.
195:
196: \begin{figure}[ht]
197: \centerline{\epsfxsize=2.5in\epsfbox{ahc_fig2.eps}}
198: \caption{Polarization as a function of wavelength. Solid circles are
199: measurements without AHC, triangles are measurements after 30 minutes of AHC,
200: and squares are measurements after 60 minutes of AHC.}
201: \end{figure}
202:
203: \section{Conclusions}
204: Atomic hydrogen cleaning can be used to prepare
205: high-QE GaAs photocathodes at the lower heat-cleaning temperature of
206: 450$^\circ$C. Photocathode quantum efficiencies as high as 15\% were
207: obtained when hydrogen ions were eliminated. Extended exposure to atomic
208: hydrogen was found to have no effect on the electron polarization.
209:
210: We thank Matt Poelker of Jefferson Lab for providing useful information about
211: the AHC technique used at his laboratory.
212:
213: \begin{thebibliography}{0}
214: \bibitem{tm}
215: T. Maruyama, D.-A. Luh, A. Brachmann, J.E. Clendenin, E.L. Garwin, S. Harvey,
216: R.E. Kirby, C.Y. Prescott, and R. Prepost, {\it Appl. Phys. Lett.} {\bf 82}, 4184 (2003).
217:
218: \bibitem{sinclair}
219: C. K. Sinclair, B. M. Poelker, and J. S. Price, {\it Proceedings of the 1997
220: IEEE Particle Accelerator Conference}, 12-16 May 1997, Vancouver, B.C., Canada,
221: p. 2864.
222:
223: \bibitem{pearton}
224: S.J. Pearton, J.W. Corbett, and M. Stavola, {\it Hydrogens in Crystalline
225: Semiconductors,} (Springer-Verlag, Heidelberg, 1992).
226:
227: \bibitem{ide}
228: Y. Ide and M. Yamada, {\it J. Vac. Sci. Technol.} {\bf A12}, 1858 (1994).
229:
230: \bibitem{baylac}
231: M. Baylac, {\it Proceedings of the 15th International Spin Physics Symposium,}
232: 9-14, September 2002, Upton, New York, p. 1073.
233:
234: \end{thebibliography}
235:
236: \end{document}
237: