Donor and acceptor competitions in phosphorus-doped ZnO - QSL
APPLIED PHYSICS LETTERS 88, 152116 共2006兲
Donor and acceptor competitions in phosphorus-doped ZnO F. X. Xiu, Z. Yang, L. J. Mandalapu, and J. L. Liua兲 Quantum Structures Laboratory, Department of Electrical Engineering, University of California, Riverside, California 92521
共Received 21 December 2005; accepted 1 March 2006; published online 14 April 2006兲 Phosphorus-doped ZnO films were grown by molecular-beam epitaxy with a GaP effusion cell as dopant source. Three growth regions were identified to obtain ZnO films with different conduction types. In the oxygen-extremely-rich region, phosphorus-doped ZnO films show n-type conduction with dominant donor-bound excitons 共D 0X兲 in the low-temperature photoluminescence 共PL兲 spectra. In the oxygen-rich region, a growth window was found to generate p-type ZnO films. The PL spectra show evident competitions between D 0X and acceptor-bound excitons 共A 0X兲. In the stoichiometric and Zn-rich region, ZnO films are n-type with dominant D 0X emissions. Thus, phosphorus doping is amphoteric, having the tendency to form both donors and acceptors in ZnO. © 2006 American Institute of Physics. 关DOI: 10.1063/1.2194870兴 In the progress of developing reliable and reproducible p-type ZnO for ZnO-based optoelectronic and spintronic applications,1 group V element phosphorus is one of the promising candidates of materials having good p-type electrical properties, such as high carrier concentration, high mobility, and low resistivity.2–9 In addition, rectifying I-V behavior was observed for diode employing phosphorus-doped p-type ZnO with n-type materials, further confirming the p-type conduction by phosphorus doping.10–14 However, the reliability of phosphorus p-type doping is still problematic. Furthermore, the lack of understanding of phosphorus doping mechanism hinders the progress of reliably producing phosphorus-doped p-type ZnO. In this work, we report a systematic study on electrical and optical properties of phosphorus-doped ZnO films grown by molecular-beam epitaxy with a GaP effusion cell as phosphorus dopant source. Combining room-temperature 共RT兲 Hall effect and lowtemperature 共LT兲 photoluminescence 共PL兲 measurements, the nature of phosphorus doping is explored. ¯ 2兲 The phosphorus-doped ZnO films were grown on 共011 r-plane sapphire substrates. Elemental zinc 共5N兲 was evaporated with a LT effusion cell. The oxygen 共5N兲 plasma was generated with a radio-frequency plasma source. All films were grown at 720 ° C with oxygen flow rate of 6 SCCM 共SCCM denotes cubic centimeter per minute at STP兲. The growth time varied from 3 to 8 h, resulting in different thicknesses, as shown in Table I. After growth, a postannealing process was performed at 800 ° C for 20 min to activate phosphorus dopants. The detailed growth procedures are presented elsewhere.8 In this study, two sets of samples were prepared, as shown in Table I. In the first set of films 共samples a–e兲, Zn cell temperatures were varied from 320 to 380 ° C while the GaP cell temperature was fixed at 710 ° C. The growth rate is plotted as a function of Zn cell temperatures in Fig. 1; three growth regions 共I–III兲 were identified accordingly. In region I, the growth rate is extremely low due to insufficient Zn atoms; therefore this region is assigned as an oxygen-extremely-rich region. RT Hall effect measurements show that the films in this region have a兲
Author to whom correspondence should be addressed; electronic mail: [email protected]
strong n-type conduction with high electron concentrations of 9.6⫻ 1018 – 2.0⫻ 1019 cm−3. In region II, the growth rate increases rapidly with Zn cell temperatures. Since the growth rate is still limited by Zn cell temperatures, this region is assigned as an oxygen-rich condition. The Hall effect measurements yield ambiguous carrier type for the films grown with Zn cell temperatures of 340 and 360 ° C 共samples b and d兲. This result may arise from the strong compensation effect with equivalent concentrations of electrons and holes in these two samples. By mediating the Zn cell temperature as 350 ° C 共sample c兲, however, strong p-type conduction was obtained with a hole concentration of 1.2⫻ 1018 cm−3 and a mobility of 4.2 cm2 / V s.8 Further increasing the Zn cell temperature beyond 360 ° C causes the saturation of growth rate, which corresponds to the stoichiometric and Zn-rich condition 共region III兲. The films in this region show n-type conduction with typical electron concentration of 2.3⫻ 1018 cm−3. To further study the properties of p-type phosphorusdoped ZnO and achieve controllable hole concentrations, the second set of films was prepared by varying GaP cell temperatures from 680, 710, to 750 ° C while keeping the Zn cell temperature at 350 ° C 共region II, samples f, c, and g, respectively兲. RT Hall effect measurements show that with GaP cell temperatures of 710 and 750 ° C, the phosphorus-doped ZnO films exhibit strong p-type conduction. For example, the sample grown with GaP cell temperature of 750 ° C 共sample g兲 has the highest hole concentration of 6.0⫻ 1018 cm−3 and a mobility of 1.5 cm2 / V s. But the film with GaP cell temperature of 680 ° C 共sample f兲 shows n-type conduction with an electron concentration of 7.8⫻ 1017 cm−3. These data suggest that an oxygen-rich condition and a high GaP cell temperature are necessary for producing p-type ZnO by phosphorus doping using a GaP effusion cell. Second ion mass spectroscopy 共SIMS兲 was performed on phosphorus-doped p-type ZnO films to examine the incorporation and concentration of phosphorus dopants using a Cameca 4f SIMS system. The SIMS depth profiles are shown in Fig. 2. For the GaP cell temperatures of 710 and 750 ° C, phosphorus doping concentrations were obtained as 2.0⫻ 1018 and 8.0⫻ 1018 cm−3, respectively. Comparing these values with the hole concentrations of 1.2⫻ 1018 and 6.0⫻ 1018 cm−3 obtained from Hall effect measurements, it is
0003-6951/2006/88共15兲/152116/3/$23.00 88, 152116-1 © 2006 American Institute of Physics Downloaded 25 Apr 2006 to 138.23.213.183. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
152116-2
Appl. Phys. Lett. 88, 152116 共2006兲
Xiu et al. TABLE I. Growth conditions and RT electrical properties of phosphorus-doped ZnO films.
Sample
Zn cell temperature 共°C兲
GaP cell temperature 共°C兲
Thickness 共m兲
Conduction type
Carrier concentration 共cm−3兲
Mobility 共cm2 / V s兲
a b c d e f g
320 340 350 360 370 350 350
710 710 710 710 710 680 750
0.29 0.53 0.40 0.55 0.62 0.40 0.74
n Ambiguous p Ambiguous n n p
2.0⫻ 1019 N/A 1.2⫻ 1018 N/A 2.3⫻ 1018 7.8⫻ 1017 6.0⫻ 1018
31.3 N/A 4.2 N/A 45.2 2.4 1.5
believed that phosphorus dopants were completely activated at the annealing temperature of 800 ° C. LT PL measurements were carried out to investigate the optical transitions in these films. A 325 nm He–Cd laser was used as the excitation source. Figures 3共a兲–3共e兲 show the PL spectra at 8.5 K for the first set of phosphorus-doped ZnO films grown with different Zn cell temperatures: 共a兲 TZn = 320 ° C 共region I兲; 关共b兲–共d兲兴 TZn = 340, 350, and 360 ° C, respectively 共region II兲; and 共e兲 TZn = 370 ° C 共region III兲. In the oxygen-extremely-rich condition 关region I, Fig. 3共a兲兴, the PL spectrum is dominated by a strong D 0X emission, which was widely observed in undoped,1 nitrogen-doped,15 phosphorus-doped,3 and arsenic-doped16 ZnO films. Another emission is observed at 3.256 eV, which may be attributed to phosphorus doping. By increasing Zn cell temperature to 340 ° C 关region II, Fig. 3共b兲兴, the features of PL spectrum remain nearly the same except for a stronger emission at 3.267 eV. The RT Hall effect measurements reveal an indeterminate carrier type for this sample. By increasing the Zn cell temperature to 350 ° C 关region II, Fig. 3共c兲兴, however, D 0X and A 0X emissions simultaneously show up in the PL spectrum at 3.364 and 3.315 eV, respectively, with a relatively stronger A 0X than D 0X. Consistently, the Hall effect measurement shows strong p-type conduction for this sample. The coexistence of D 0X and A 0X in this spectrum indicates that there are competitions between donors and acceptors in this film. By further increasing the Zn cell temperature to 360 ° C 关region II, Fig. 3共d兲兴, the spectrum shows a lower intensity of A 0X than D 0X. Correspondingly, the Hall effect measurement does not yield a clear p-type conduction, similar to the scenario of TZn = 340 ° C 关region II,
Fig. 3共b兲兴. The emissions at 3.259 and 3.180 eV were clearly observed and identified as free electron to acceptor 共FA兲 and donor-acceptor pair 共DAP兲 transitions, respectively, in our previous study.8 Finally, raising the Zn cell temperature to 370 ° C leads to a dominant D 0X emission at 3.361 eV and a phosphorus-related emission at 3.303 eV 关region III, Fig. 3共e兲兴. The Hall effect measurement reveals n-type conduction for this sample. From the evolution of PL spectra and Hall data, it is concluded that there are competitions between D 0X and A 0X in phosphorus-doped ZnO films, and by varying the Zn cell temperatures, the intensities of D 0X and A 0X emissions can be well controlled. The final carrier type is determined by the competition results. Figures 3共f兲, 3共c兲, and 3共g兲 show the PL spectra at 8.5 K for the second set of phosphorus-doped ZnO films grown with different GaP cell temperatures of 680, 710, and 750 ° C, respectively. With a low GaP cell temperature of 680 ° C 关Fig. 3共f兲兴, the film shows n-type conduction with a D 0X dominated PL spectrum, similar to the case in Fig. 3共e兲. By increasing GaP cell temperature to 710 ° C 关Fig. 3共c兲兴, the film becomes p-type and the spectrum shows the coexistence of D 0X and A 0X with a stronger A 0X emission, as analyzed above. Further raising the GaP cell temperature to 750 ° C 关Fig. 3共g兲兴 results in a completely A 0X dominated PL spectrum at 3.317 eV and a nearly-diminished D 0X at 3.360 eV. Accordingly, the Hall effect measurement shows much stronger p-type conduction with a high hole concentration of 6.0⫻ 1018 cm−3 and a mobility of 1.5 cm2 / V s. The dramatic change of PL spectra in the second set of samples further proves the existence of competitions between D 0X and A 0X in phosphorus-doped ZnO films. The doping mechanism for phosphorus-doped ZnO remains unclear and sometimes controversial.4,6,7,17–24 Recent
FIG. 1. Phosphorus-doped ZnO growth rate as a function of Zn cell temperature. Three regions were identified: Region I, oxygen-extremely-rich region 共n type兲; region II, oxygen-rich region 共p type兲; and region III, FIG. 2. The SIMS depth profiles for phosphorus-doped p-type ZnO films stoichiometric and Zn-rich region 共n type兲. Samples a–e were also marked in with TGaP = 710 and 750 ° C. Phosphorus doping concentrations are obtained as 2.0⫻ 1018 and 8.0⫻ 1018 cm−3, respectively. this figure. The growth conditions can be found in Table I. Downloaded 25 Apr 2006 to 138.23.213.183. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
152116-3
Appl. Phys. Lett. 88, 152116 共2006兲
Xiu et al.
consistency of these experimental results confirmed that the phosphorus doping is of amphoteric character. In summary, phosphorus-doped ZnO films were produced using a GaP cell as the phosphorus dopant source in a molecular-beam epitaxy system. Three growth regions were identified to generate ZnO films with different conduction types by systematically varying the Zn cell temperatures. A growth window was found to generate p-type ZnO films. The PL spectra clearly indicate the existence of competitions between D 0X and A 0X for the phosphorus-doped ZnO films, especially in the oxygen-rich region. This study suggests that to achieve reliable p-type ZnO by phosphorus doping, an adoption of oxygen-rich growth condition and high phosphorus incorporation during growth is needed to suppress/ compensate the formation of donors. Hence, a high acceptor concentration could be achieved. Phosphorus doping mechanism is confirmed to be amphoteric. Depending on growth conditions, phosphorus-related donors and acceptors can be formed with different concentrations. The final carrier type is determined by the competition results. This work was supported by DARPA/DMEA through the center for NanoScience and Innovation for Defense 共CNID兲 under Award No. H94003-04-2-0404. D. C. Look and B. Claflin, Phys. Status Solidi B 241, 624 共2004兲. K. K. Kim, H. S. Kim, D. K. Hwang, J. H. Lim, and S. J. Park, Appl. Phys. Lett. 83, 63 共2003兲. 3 D. K. Hwang, H. S. Kim, J. H. Lim, J. Y. Oh, J. H. Yang, S. J. Park, K. K. Kim, D. C. Look, and Y. S. Park, Appl. Phys. Lett. 86, 151917 共2005兲. 4 F. G. Chen, Z. Z. Ye, W. Z. Xu, B. H. Zhao, L. P. Zhu, and J. G. Lv, J. Cryst. Growth 281, 458 共2005兲. 5 V. Vaithianathan, B. T. Lee, and S. S. Kim, J. Appl. Phys. 98, 043519 共2005兲. FIG. 3. PL spectra at 8.5 K for two sets of phosphorus-doped ZnO films. 6 Y. J. Li, Y. W. Heo, Y. Kwon, K. Ip, S. J. Pearton, and D. P. Norton, 共a兲–共g兲 are corresponding to samples a–g, as shown in Table I. Appl. Phys. Lett. 87, 072101 共2005兲. 7 S. J. So and C. B. Park, J. Cryst. Growth 285, 606 共2005兲. 8 studies showed that the incorporation of phosphorus leads to F. X. Xiu, Z. Yang, L. J. Mandalapu, J. L. Liu, and W. P. Beyermann, Appl. Phys. Lett. 88, 052106 共2006兲. a significant increase of electron concentration,5,18,20,21 sug9 Y. Miao, Z. Z. Ye, W. Z. Xu, F. G. Chen, X. C. Zhou, B. G. Zhao, L. P. gesting that the formation of anti-site defects PZn and P+3-, Zhu, and J. G. Lu, Appl. Surf. Sci. 共unpublished兲. +5 −3 P -, or P -related complex could be the origin of shallow 10 H. Yang, Y. Li, D. P. Norton, S. J. Pearton, S. Jung, F. Ren, and L. A. 20–22 Consistently, donors in phosphorus-doped n-type ZnO. Boatner, Appl. Phys. Lett. 86, 172103 共2005兲. 11 our results show that, in the oxygen-extremely-rich condiT. Aoki, Y. Hatanaka, and D. C. Look, Appl. Phys. Lett. 76, 3257 共2000兲. 12 K. H. Bang, D. K. Hwang, M. C. Park, Y. D. Ko, I. Yun, and J. M. tion, the phosphorus-doped ZnO films become very conducMyoung, Appl. Surf. Sci. 210, 177 共2003兲. tive with electron concentrations up to 2.0⫻ 1019 cm−3. In 13 Y. D. Ko, J. Jung, K. H. Bang, M. C. Park, K. S. Huh, J. M. Myoung, and the oxygen-rich condition, however, phosphorus-related A 0X I. Yun, Appl. Surf. Sci. 202, 266 共2002兲. 14 emission starts to show up 关Fig. 3共c兲兴 and becomes much S. Jang, J. J. Chen, B. S. Kang, F. Ren, D. P. Norton, S. J. Pearton, J. stronger by raising GaP cell temperature to 750 ° C 关Fig. Lopata, and W. S. Hobson, Appl. Phys. Lett. 87, 222113 共2005兲. 15 D. C. Look, D. C. Reynolds, C. W. Litton, R. L. Jones, D. B. Eason, and 3共g兲兴, which clearly indicates that high-density acceptors G. Cantwell, Appl. Phys. Lett. 81, 1830 共2002兲. were incorporated into the films. Theoretical work has been 16 Y. R. Ryu, T. S. Lee, and H. W. White, Appl. Phys. Lett. 83, 87 共2003兲. done to understand the origin of p-type conduction by phos17 H. Tampo, H. Shibata, P. Fons, A. Yamada, K. Matsubara, K. Iwata, K. 22–24 It was suggested that a possible shallow phorus doping. Tamura, H. Takasu, and S. Niki, J. Cryst. Growth 278, 268 共2005兲. −1 18 acceptor PO could be formed with an energy level of Z. Q. Chen, A. Kawasuso, Y. Xu, H. Naramoto, X. L. Yuan, T. Sekiguchi, 22 0.49 eV above the valence band maximum. ExperimenR. Suzuki, and T. Ohdaira, J. Appl. Phys. 97, 013528 共2005兲. 19 Y. W. Heo, Y. W. Kwon, Y. Li, S. J. Pearton, and D. P. Norton, Appl. Phys. tally, similar to As- and Sb-doped ZnO,3,25,26 the acceptor Lett. 84, 3474 共2004兲. activation energy of phosphorus-doped ZnO films was 20 Y. W. Heo, K. Ip, S. J. Park, S. J. Pearton, and D. P. Norton, Appl. Phys. shown to be much smaller than theoretical calculations, in A: Mater. Sci. Process. 78, 53 共2004兲. 3,8 21 the range of 127– 180 meV. Nevertheless, the high hole Y. W. Heo, S. J. Park, K. Ip, S. J. Pearton, and D. P. Norton, Appl. Phys. concentration obtained in the present study indicates the forLett. 83, 1128 共2003兲. 22 Z. G. Yu, H. Gong, and P. Wu, Appl. Phys. Lett. 86, 212105 共2005兲. mation of shallow phosphorus-related acceptors. It is also 23 Z. G. Yu, H. Gong, and P. Wu, Chem. Mater. 17, 852 共2005兲. interesting to note that similar experimental results were ob24 C. H. Park, S. B. Zhang, and S. H. Wei, Phys. Rev. B 66, 073202 共2002兲. tained for pulsed laser deposition 共PLD兲-grown ZnO films 25 F. X. Xiu, Z. Yang, L. J. Mandalapu, D. T. Zhao, J. L. Liu, and W. P. 6 using P2O5 as dopant source. Oxygen flow rates were varied Beyermann, Appl. Phys. Lett. 87, 152101 共2005兲. 26 from 20 to 200 mTorr and a narrow p-type growth window F. X. Xiu, Z. Yang, L. J. Mandalapu, D. T. Zhao, and J. L. Liu, Appl. Phys. was found only with an oxygen flow rate of 150 mTorr. The Lett. 87, 252102 共2005兲. Downloaded 25 Apr 2006 to 138.23.213.183. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp 1 2
Donor and acceptor competitions in phosphorus-doped ZnO F. X. Xiu, Z. Yang, L. J. Mandalapu, and J. L. Liua兲 Quantum Structures Laboratory, Department of Electrical Engineering, University of California, Riverside, California 92521
共Received 21 December 2005; accepted 1 March 2006; published online 14 April 2006兲 Phosphorus-doped ZnO films were grown by molecular-beam epitaxy with a GaP effusion cell as dopant source. Three growth regions were identified to obtain ZnO films with different conduction types. In the oxygen-extremely-rich region, phosphorus-doped ZnO films show n-type conduction with dominant donor-bound excitons 共D 0X兲 in the low-temperature photoluminescence 共PL兲 spectra. In the oxygen-rich region, a growth window was found to generate p-type ZnO films. The PL spectra show evident competitions between D 0X and acceptor-bound excitons 共A 0X兲. In the stoichiometric and Zn-rich region, ZnO films are n-type with dominant D 0X emissions. Thus, phosphorus doping is amphoteric, having the tendency to form both donors and acceptors in ZnO. © 2006 American Institute of Physics. 关DOI: 10.1063/1.2194870兴 In the progress of developing reliable and reproducible p-type ZnO for ZnO-based optoelectronic and spintronic applications,1 group V element phosphorus is one of the promising candidates of materials having good p-type electrical properties, such as high carrier concentration, high mobility, and low resistivity.2–9 In addition, rectifying I-V behavior was observed for diode employing phosphorus-doped p-type ZnO with n-type materials, further confirming the p-type conduction by phosphorus doping.10–14 However, the reliability of phosphorus p-type doping is still problematic. Furthermore, the lack of understanding of phosphorus doping mechanism hinders the progress of reliably producing phosphorus-doped p-type ZnO. In this work, we report a systematic study on electrical and optical properties of phosphorus-doped ZnO films grown by molecular-beam epitaxy with a GaP effusion cell as phosphorus dopant source. Combining room-temperature 共RT兲 Hall effect and lowtemperature 共LT兲 photoluminescence 共PL兲 measurements, the nature of phosphorus doping is explored. ¯ 2兲 The phosphorus-doped ZnO films were grown on 共011 r-plane sapphire substrates. Elemental zinc 共5N兲 was evaporated with a LT effusion cell. The oxygen 共5N兲 plasma was generated with a radio-frequency plasma source. All films were grown at 720 ° C with oxygen flow rate of 6 SCCM 共SCCM denotes cubic centimeter per minute at STP兲. The growth time varied from 3 to 8 h, resulting in different thicknesses, as shown in Table I. After growth, a postannealing process was performed at 800 ° C for 20 min to activate phosphorus dopants. The detailed growth procedures are presented elsewhere.8 In this study, two sets of samples were prepared, as shown in Table I. In the first set of films 共samples a–e兲, Zn cell temperatures were varied from 320 to 380 ° C while the GaP cell temperature was fixed at 710 ° C. The growth rate is plotted as a function of Zn cell temperatures in Fig. 1; three growth regions 共I–III兲 were identified accordingly. In region I, the growth rate is extremely low due to insufficient Zn atoms; therefore this region is assigned as an oxygen-extremely-rich region. RT Hall effect measurements show that the films in this region have a兲
Author to whom correspondence should be addressed; electronic mail: [email protected]
strong n-type conduction with high electron concentrations of 9.6⫻ 1018 – 2.0⫻ 1019 cm−3. In region II, the growth rate increases rapidly with Zn cell temperatures. Since the growth rate is still limited by Zn cell temperatures, this region is assigned as an oxygen-rich condition. The Hall effect measurements yield ambiguous carrier type for the films grown with Zn cell temperatures of 340 and 360 ° C 共samples b and d兲. This result may arise from the strong compensation effect with equivalent concentrations of electrons and holes in these two samples. By mediating the Zn cell temperature as 350 ° C 共sample c兲, however, strong p-type conduction was obtained with a hole concentration of 1.2⫻ 1018 cm−3 and a mobility of 4.2 cm2 / V s.8 Further increasing the Zn cell temperature beyond 360 ° C causes the saturation of growth rate, which corresponds to the stoichiometric and Zn-rich condition 共region III兲. The films in this region show n-type conduction with typical electron concentration of 2.3⫻ 1018 cm−3. To further study the properties of p-type phosphorusdoped ZnO and achieve controllable hole concentrations, the second set of films was prepared by varying GaP cell temperatures from 680, 710, to 750 ° C while keeping the Zn cell temperature at 350 ° C 共region II, samples f, c, and g, respectively兲. RT Hall effect measurements show that with GaP cell temperatures of 710 and 750 ° C, the phosphorus-doped ZnO films exhibit strong p-type conduction. For example, the sample grown with GaP cell temperature of 750 ° C 共sample g兲 has the highest hole concentration of 6.0⫻ 1018 cm−3 and a mobility of 1.5 cm2 / V s. But the film with GaP cell temperature of 680 ° C 共sample f兲 shows n-type conduction with an electron concentration of 7.8⫻ 1017 cm−3. These data suggest that an oxygen-rich condition and a high GaP cell temperature are necessary for producing p-type ZnO by phosphorus doping using a GaP effusion cell. Second ion mass spectroscopy 共SIMS兲 was performed on phosphorus-doped p-type ZnO films to examine the incorporation and concentration of phosphorus dopants using a Cameca 4f SIMS system. The SIMS depth profiles are shown in Fig. 2. For the GaP cell temperatures of 710 and 750 ° C, phosphorus doping concentrations were obtained as 2.0⫻ 1018 and 8.0⫻ 1018 cm−3, respectively. Comparing these values with the hole concentrations of 1.2⫻ 1018 and 6.0⫻ 1018 cm−3 obtained from Hall effect measurements, it is
0003-6951/2006/88共15兲/152116/3/$23.00 88, 152116-1 © 2006 American Institute of Physics Downloaded 25 Apr 2006 to 138.23.213.183. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
152116-2
Appl. Phys. Lett. 88, 152116 共2006兲
Xiu et al. TABLE I. Growth conditions and RT electrical properties of phosphorus-doped ZnO films.
Sample
Zn cell temperature 共°C兲
GaP cell temperature 共°C兲
Thickness 共m兲
Conduction type
Carrier concentration 共cm−3兲
Mobility 共cm2 / V s兲
a b c d e f g
320 340 350 360 370 350 350
710 710 710 710 710 680 750
0.29 0.53 0.40 0.55 0.62 0.40 0.74
n Ambiguous p Ambiguous n n p
2.0⫻ 1019 N/A 1.2⫻ 1018 N/A 2.3⫻ 1018 7.8⫻ 1017 6.0⫻ 1018
31.3 N/A 4.2 N/A 45.2 2.4 1.5
believed that phosphorus dopants were completely activated at the annealing temperature of 800 ° C. LT PL measurements were carried out to investigate the optical transitions in these films. A 325 nm He–Cd laser was used as the excitation source. Figures 3共a兲–3共e兲 show the PL spectra at 8.5 K for the first set of phosphorus-doped ZnO films grown with different Zn cell temperatures: 共a兲 TZn = 320 ° C 共region I兲; 关共b兲–共d兲兴 TZn = 340, 350, and 360 ° C, respectively 共region II兲; and 共e兲 TZn = 370 ° C 共region III兲. In the oxygen-extremely-rich condition 关region I, Fig. 3共a兲兴, the PL spectrum is dominated by a strong D 0X emission, which was widely observed in undoped,1 nitrogen-doped,15 phosphorus-doped,3 and arsenic-doped16 ZnO films. Another emission is observed at 3.256 eV, which may be attributed to phosphorus doping. By increasing Zn cell temperature to 340 ° C 关region II, Fig. 3共b兲兴, the features of PL spectrum remain nearly the same except for a stronger emission at 3.267 eV. The RT Hall effect measurements reveal an indeterminate carrier type for this sample. By increasing the Zn cell temperature to 350 ° C 关region II, Fig. 3共c兲兴, however, D 0X and A 0X emissions simultaneously show up in the PL spectrum at 3.364 and 3.315 eV, respectively, with a relatively stronger A 0X than D 0X. Consistently, the Hall effect measurement shows strong p-type conduction for this sample. The coexistence of D 0X and A 0X in this spectrum indicates that there are competitions between donors and acceptors in this film. By further increasing the Zn cell temperature to 360 ° C 关region II, Fig. 3共d兲兴, the spectrum shows a lower intensity of A 0X than D 0X. Correspondingly, the Hall effect measurement does not yield a clear p-type conduction, similar to the scenario of TZn = 340 ° C 关region II,
Fig. 3共b兲兴. The emissions at 3.259 and 3.180 eV were clearly observed and identified as free electron to acceptor 共FA兲 and donor-acceptor pair 共DAP兲 transitions, respectively, in our previous study.8 Finally, raising the Zn cell temperature to 370 ° C leads to a dominant D 0X emission at 3.361 eV and a phosphorus-related emission at 3.303 eV 关region III, Fig. 3共e兲兴. The Hall effect measurement reveals n-type conduction for this sample. From the evolution of PL spectra and Hall data, it is concluded that there are competitions between D 0X and A 0X in phosphorus-doped ZnO films, and by varying the Zn cell temperatures, the intensities of D 0X and A 0X emissions can be well controlled. The final carrier type is determined by the competition results. Figures 3共f兲, 3共c兲, and 3共g兲 show the PL spectra at 8.5 K for the second set of phosphorus-doped ZnO films grown with different GaP cell temperatures of 680, 710, and 750 ° C, respectively. With a low GaP cell temperature of 680 ° C 关Fig. 3共f兲兴, the film shows n-type conduction with a D 0X dominated PL spectrum, similar to the case in Fig. 3共e兲. By increasing GaP cell temperature to 710 ° C 关Fig. 3共c兲兴, the film becomes p-type and the spectrum shows the coexistence of D 0X and A 0X with a stronger A 0X emission, as analyzed above. Further raising the GaP cell temperature to 750 ° C 关Fig. 3共g兲兴 results in a completely A 0X dominated PL spectrum at 3.317 eV and a nearly-diminished D 0X at 3.360 eV. Accordingly, the Hall effect measurement shows much stronger p-type conduction with a high hole concentration of 6.0⫻ 1018 cm−3 and a mobility of 1.5 cm2 / V s. The dramatic change of PL spectra in the second set of samples further proves the existence of competitions between D 0X and A 0X in phosphorus-doped ZnO films. The doping mechanism for phosphorus-doped ZnO remains unclear and sometimes controversial.4,6,7,17–24 Recent
FIG. 1. Phosphorus-doped ZnO growth rate as a function of Zn cell temperature. Three regions were identified: Region I, oxygen-extremely-rich region 共n type兲; region II, oxygen-rich region 共p type兲; and region III, FIG. 2. The SIMS depth profiles for phosphorus-doped p-type ZnO films stoichiometric and Zn-rich region 共n type兲. Samples a–e were also marked in with TGaP = 710 and 750 ° C. Phosphorus doping concentrations are obtained as 2.0⫻ 1018 and 8.0⫻ 1018 cm−3, respectively. this figure. The growth conditions can be found in Table I. Downloaded 25 Apr 2006 to 138.23.213.183. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
152116-3
Appl. Phys. Lett. 88, 152116 共2006兲
Xiu et al.
consistency of these experimental results confirmed that the phosphorus doping is of amphoteric character. In summary, phosphorus-doped ZnO films were produced using a GaP cell as the phosphorus dopant source in a molecular-beam epitaxy system. Three growth regions were identified to generate ZnO films with different conduction types by systematically varying the Zn cell temperatures. A growth window was found to generate p-type ZnO films. The PL spectra clearly indicate the existence of competitions between D 0X and A 0X for the phosphorus-doped ZnO films, especially in the oxygen-rich region. This study suggests that to achieve reliable p-type ZnO by phosphorus doping, an adoption of oxygen-rich growth condition and high phosphorus incorporation during growth is needed to suppress/ compensate the formation of donors. Hence, a high acceptor concentration could be achieved. Phosphorus doping mechanism is confirmed to be amphoteric. Depending on growth conditions, phosphorus-related donors and acceptors can be formed with different concentrations. The final carrier type is determined by the competition results. This work was supported by DARPA/DMEA through the center for NanoScience and Innovation for Defense 共CNID兲 under Award No. H94003-04-2-0404. D. C. Look and B. Claflin, Phys. Status Solidi B 241, 624 共2004兲. K. K. Kim, H. S. Kim, D. K. Hwang, J. H. Lim, and S. J. Park, Appl. Phys. Lett. 83, 63 共2003兲. 3 D. K. Hwang, H. S. Kim, J. H. Lim, J. Y. Oh, J. H. Yang, S. J. Park, K. K. Kim, D. C. Look, and Y. S. Park, Appl. Phys. Lett. 86, 151917 共2005兲. 4 F. G. Chen, Z. Z. Ye, W. Z. Xu, B. H. Zhao, L. P. Zhu, and J. G. Lv, J. Cryst. Growth 281, 458 共2005兲. 5 V. Vaithianathan, B. T. Lee, and S. S. Kim, J. Appl. Phys. 98, 043519 共2005兲. FIG. 3. PL spectra at 8.5 K for two sets of phosphorus-doped ZnO films. 6 Y. J. Li, Y. W. Heo, Y. Kwon, K. Ip, S. J. Pearton, and D. P. Norton, 共a兲–共g兲 are corresponding to samples a–g, as shown in Table I. Appl. Phys. Lett. 87, 072101 共2005兲. 7 S. J. So and C. B. Park, J. Cryst. Growth 285, 606 共2005兲. 8 studies showed that the incorporation of phosphorus leads to F. X. Xiu, Z. Yang, L. J. Mandalapu, J. L. Liu, and W. P. Beyermann, Appl. Phys. Lett. 88, 052106 共2006兲. a significant increase of electron concentration,5,18,20,21 sug9 Y. Miao, Z. Z. Ye, W. Z. Xu, F. G. Chen, X. C. Zhou, B. G. Zhao, L. P. gesting that the formation of anti-site defects PZn and P+3-, Zhu, and J. G. Lu, Appl. Surf. Sci. 共unpublished兲. +5 −3 P -, or P -related complex could be the origin of shallow 10 H. Yang, Y. Li, D. P. Norton, S. J. Pearton, S. Jung, F. Ren, and L. A. 20–22 Consistently, donors in phosphorus-doped n-type ZnO. Boatner, Appl. Phys. Lett. 86, 172103 共2005兲. 11 our results show that, in the oxygen-extremely-rich condiT. Aoki, Y. Hatanaka, and D. C. Look, Appl. Phys. Lett. 76, 3257 共2000兲. 12 K. H. Bang, D. K. Hwang, M. C. Park, Y. D. Ko, I. Yun, and J. M. tion, the phosphorus-doped ZnO films become very conducMyoung, Appl. Surf. Sci. 210, 177 共2003兲. tive with electron concentrations up to 2.0⫻ 1019 cm−3. In 13 Y. D. Ko, J. Jung, K. H. Bang, M. C. Park, K. S. Huh, J. M. Myoung, and the oxygen-rich condition, however, phosphorus-related A 0X I. Yun, Appl. Surf. Sci. 202, 266 共2002兲. 14 emission starts to show up 关Fig. 3共c兲兴 and becomes much S. Jang, J. J. Chen, B. S. Kang, F. Ren, D. P. Norton, S. J. Pearton, J. stronger by raising GaP cell temperature to 750 ° C 关Fig. Lopata, and W. S. Hobson, Appl. Phys. Lett. 87, 222113 共2005兲. 15 D. C. Look, D. C. Reynolds, C. W. Litton, R. L. Jones, D. B. Eason, and 3共g兲兴, which clearly indicates that high-density acceptors G. Cantwell, Appl. Phys. Lett. 81, 1830 共2002兲. were incorporated into the films. Theoretical work has been 16 Y. R. Ryu, T. S. Lee, and H. W. White, Appl. Phys. Lett. 83, 87 共2003兲. done to understand the origin of p-type conduction by phos17 H. Tampo, H. Shibata, P. Fons, A. Yamada, K. Matsubara, K. Iwata, K. 22–24 It was suggested that a possible shallow phorus doping. Tamura, H. Takasu, and S. Niki, J. Cryst. Growth 278, 268 共2005兲. −1 18 acceptor PO could be formed with an energy level of Z. Q. Chen, A. Kawasuso, Y. Xu, H. Naramoto, X. L. Yuan, T. Sekiguchi, 22 0.49 eV above the valence band maximum. ExperimenR. Suzuki, and T. Ohdaira, J. Appl. Phys. 97, 013528 共2005兲. 19 Y. W. Heo, Y. W. Kwon, Y. Li, S. J. Pearton, and D. P. Norton, Appl. Phys. tally, similar to As- and Sb-doped ZnO,3,25,26 the acceptor Lett. 84, 3474 共2004兲. activation energy of phosphorus-doped ZnO films was 20 Y. W. Heo, K. Ip, S. J. Park, S. J. Pearton, and D. P. Norton, Appl. Phys. shown to be much smaller than theoretical calculations, in A: Mater. Sci. Process. 78, 53 共2004兲. 3,8 21 the range of 127– 180 meV. Nevertheless, the high hole Y. W. Heo, S. J. Park, K. Ip, S. J. Pearton, and D. P. Norton, Appl. Phys. concentration obtained in the present study indicates the forLett. 83, 1128 共2003兲. 22 Z. G. Yu, H. Gong, and P. Wu, Appl. Phys. Lett. 86, 212105 共2005兲. mation of shallow phosphorus-related acceptors. It is also 23 Z. G. Yu, H. Gong, and P. Wu, Chem. Mater. 17, 852 共2005兲. interesting to note that similar experimental results were ob24 C. H. Park, S. B. Zhang, and S. H. Wei, Phys. Rev. B 66, 073202 共2002兲. tained for pulsed laser deposition 共PLD兲-grown ZnO films 25 F. X. Xiu, Z. Yang, L. J. Mandalapu, D. T. Zhao, J. L. Liu, and W. P. 6 using P2O5 as dopant source. Oxygen flow rates were varied Beyermann, Appl. Phys. Lett. 87, 152101 共2005兲. 26 from 20 to 200 mTorr and a narrow p-type growth window F. X. Xiu, Z. Yang, L. J. Mandalapu, D. T. Zhao, and J. L. Liu, Appl. Phys. was found only with an oxygen flow rate of 150 mTorr. The Lett. 87, 252102 共2005兲. Downloaded 25 Apr 2006 to 138.23.213.183. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp 1 2
