Growth of the Dilute Magnetic Semiconductor GaMnN by Molecular Beam Epitaxy

M. E.Overberg^1,*, G. T. Thaler¹, C. R. Abernathy¹, N. A. Theodoropoulou², K. T. McCarthy², S. B. Arnason², J. S.Lee³, J. D.Lim³, S. B.Shim³, K. S.Suh³, Z. G.Khim³, Y. D.Park³, S. J.Pearton¹, and A. F. Hebard²

¹Department of Materials Science and Engineering, University of Florida, Gainesville, Florida, 32611

²Department of Physics, University of Florida, Gainesville, Florida, 32611

³ Center for Strongly Correlated Materials Research, Seoul National University, Seoul, 151-747, Seoul, Korea

ABSTRACT

Growth by molecular-beam epitaxy of the dilute magnetic alloy GaMnN is reported. The Mn concentration as determined by Auger electron spectroscopy is found to be linear with increasing Mn cell temperature up to ~43 at % Mn. No second phases are observed for Mn levels below 9 at %. The cubic phase Mn₄N is found to be the thermodynamically stable phase at the growth conditions used to produce GaMnN. Hysteresis in M versus H is observed in both GaMnN and GaMnN:C grown on both sapphire and MOCVD GaN, at several growth temperatures. Magnetotransport results show the anomalous Hall effect, negative magnetoresistance, and magnetic hysteresis, indicating that Mn is incorporating into the GaN and forming the ferromagnetic semiconductor GaMnN. Room temperature hysteresis is obtained in magnetization measurements, with an optimum Mn concentration of ~3 at.%.

INTRODUCTION

Currently, a large research effort is directed towards the creation of “spintronic” materials and devices that utilize the quantum properties of the electron spin wavefunction for sensing and switching applications.^(1-6) However, the approach of directly mating electronic materials (semiconductors) with spin materials (ferromagnetic metals) leads to interfacial problems due to the dissimilar nature of the materials’ crystal structure, bonding, physical and chemical properties.⁽⁷⁾ Another solution is the use of dilute magnetic semiconductors (DMS), which consist of semiconductor hosts heavily doped with substitutional magnetic ions. A DMS material could permit direct integration with current semiconductor devices.

Several theories have been presented on the nature of DMS-related ferromagnetism.^(8,9) In one theory based on the bound magnetic polaron (BMP) model, Curie temperatures (T_C) have been calculated for 2.5 at % Mn in various III-V and II-VI semiconductors. In this calculation, a concentration of free holes equal to 3.5x10²⁰/cm³ has been assumed.⁽⁹⁾ To date, the best experimental T_C values for InMnAs, GaMnSb, and GaMnAs are in reasonable agreement with theory, but are still well below room temperature.^(10-12) The p-type III-V DMS material GaMnN is predicted to have a T_C well above room temperature.⁽⁹⁾ GaMnN has been produced in powder and crystallite form, but to this point its application has been limited.^(13,14) In addition, ferromagnetism has been demonstrated in Mn-implanted p-GaN epitaxial layers⁽¹⁵⁾ , Mn-diffused GaN ^(16,17) and in GaN grown by Molecular Beam Epitaxy^(18-22).

In the epitaxy of the III-Mn-As and III-Mn-Sb, reduced growth temperatures were used to incorporate the necessary 2.5 at % of Mn and to avoid the formation of MnAs or MnSb precipitates.^(10-12) For example, epitaxial GaAs is normally grown at 600 °C, while the GaMnAs was grown at 250 °C. For the epitaxial growth of GaMnN, the growth temperature cannot be dropped to such an extent. Amorphous GaN will begin to form if T_growth is less than approximately 600 °C. However, the relatively high vapor pressure of Mn at normal GaN growth temperatures (T_growth = 1000 °C) may make an incorporation of 2.5 at % Mn difficult to achieve. This is believed to be the limiting factor in producing GaMnN. In this paper, we demonstrate the thin film growth of ferromagnetic GaMnN, with emphasis placed on the conditions necessary to incorporate the large fraction of Mn while avoiding the formation of second phases within the material.

EXPERIMENTAL PROCEDURE

The GaMnN films were grown by Gas Source Molecular Beam Epitaxy in an INTEVAC Gas Source Gen II on In-mounted (0001) Al₂O₃ substrates and on MOCVD derived GaN surfaces. For the films grown on sapphire, a surface nitridation step was performed by exposing the Al₂O₃ to a nitrogen plasma for 5 minutes at a substrate temperature of 865°C. After the nitridation, a low-temperature AlN buffer layer was grown at a substrate temperature of 435°C for 5 minutes, resulting in 20 nm of material. Following the AlN growth, a GaN layer was grown, followed by the GaMnN layer. For the material grown directly on the MOCVD GaN surface, an oxide desorption step was performed for 10 minutes at 750 °C under no plasma, allowing the RHEED pattern to change from amorphous to 2D. The GaMnN layer was then grown. The Al for the low temperature buffer layer was provided by a dimethylethylamine alane source. Shuttered effusion ovens charged with 7 N (99.99999 % pure) Ga and 4 N (99.99 % pure) Mn provided the group III and the magnetic dopant fluxes. Reactive nitrogen for all the growth steps was provided by an SVT radio-frequency (RF) plasma source operating at 375 W of forward power and a gas flow rate of 3 sccm N₂. X-ray diffraction (XRD) measurements were performed in a Philips APD 3720 powder diffractometer for q-2q analysis, while w-2q rocking curves were measured in a Philips X’pert MPD diffractometer. Compositional information was provided by Auger Electron Spectroscopy (AES) in a Perkin-Elmer PHI 6600 system. Magnetic measurements were performed in a Quantum Design superconducting quantum interference device (SQUID) Magnetic Properties Measurement System and a Quantum Design Physical Properties Measurement System with a Linear Research LR700 AC impedance bridge. Hall measurements were also taken at room temperature in a custom built system using a 0.8 T electromagnet. Crystal quality was measured in-situ using a Staib reflection high-energy electron diffraction (RHEED) gun set to 15 kV.

RESULTS AND DISCUSSION

Incorporation of Mn into GaN

During the initial investigation of GaMnN on sapphire, the GaN layer was grown for 40 minutes at a temperature of 750 °C, while the GaMnN layer was grown for 20 minutes, also at 750 °C. The growth rate of GaN under the conditions used was approximately 69 Å/min. Room temperature Hall effect measurements indicate that the GaN is n-type with an electron concentration of 4 X 10¹⁹/cm³. The high level of donors is believed to be due to a large concentration of N vacancies, although the possibility of excess donors due to the presence of O cannot be discounted. However, SIMS analysis on other samples grown in this chamber have shown O levels much lower than 10¹⁹/cm³.

For the GaMnN growth, after the 40 minutes of GaN growth, the shutter for the 4N Mn K-cell was opened. No other system parameters were changed. A series of samples was grown using Mn cell temperatures that varied between set points of 675 °C and 937 °C, which corresponds to a change in the Mn vapor pressure of three orders of magnitude. After growth, the Mn level was evaluated using Auger electron spectroscopy (AES). For the range of cell temperatures used, the AES analysis showed the Mn incorporation to be linear with Mn cell temperature for Mn concentrations in the range of ~1.1 at % to 43.1 at %, as shown in Figure 1a. The high concentrations that could be achieved suggest that in spite of the high vapor pressure, the Mn sticking coefficient is adequate even at elevated GaN growth temperatures. The [Mn] vs. 1/T_Mn behavior was fit with an Arrhenius-type curve (shown in Figure 1a), and an activation energy for Mn incorporation was determined to be 1.53 eV. This compares to an activation energy of 2.57 eV from the [Mn] vs. 1/T_Mn vapor pressure curve. The lower activation energy would indicate easy incorporation of Mn into the epitaxially growing GaN. The growth rate of the GaMnN is plotted as a function of 1/T_Mn in Figure 1b. For the lower Mn fluxes, Figure 1b indicates a slight drop in the growth rate. This can possibly be due to the Mn slightly lowering the Ga sticking coefficient in this range of concentrations. However, as the Mn flux is increased, the GaMnN growth rate increases dramatically. This is most likely due to the Mn flux at this temperature becoming comparable to the Ga flux, producing a Mn_xGa_y alloy rather than GaMnN. This is supported by electrical measurements of the material carrier concentration. As the Mn concentration was increased in the single-phase GaN semiconductor, the GaMnN became increasingly resistive. This is in keeping with the expectation of a deep Mn acceptor (as in GaMnAs) that would serve to compensate the shallow nitrogen vacancy donor. At the very high Mn concentrations, Mn » 47 at %, the films become conductive, presumably due to formation of the MnGa phases. MnGa is a metallic ferromagnet.

The GaMnN material from this first series of growth experiments was investigated using x-ray diffraction to determine the presence of second phases within the material. Specifically, it was desired to know at what Mn level, as determined by AES, would second phases begin to appear. For the successful growth of GaMnN at a temperature of 750 °C, the Mn level must therefore be kept below this level. Magnetic measurements of GaMnN with higher concentrations of Mn would tend to measure this second phase, and not the DMS. As with the arsenides and the antimonides, the different Mn-N phases possess some magnetism. A powder x-ray diffraction scan of GaMnN with T_Mn = 747 °C is given in Figure 2a. As can be seen from the figure, the only peaks that appear in the data are the GaN K_a (002) (34.6°), the GaN K_a (004) (72.8°), the sapphire K_a (006) (42.0°), the sapphire K_a (0012) (90.8°), and the sapphire K_b (0012) (80.1°). The remaining weak K_b peaks are lost in the noise of the measurement. This T_Mn corresponds to a Mn level of 8.8 at % as determined by AES. Further increasing the T_Mn to 887 °C produced the diffraction pattern given in Figure 2b. This produces a Mn level of 43.1 at % as determined by AES. In addition to the GaN and sapphire peaks due to the underlying material, peaks of the tetragonal Mn_0.6Ga_0.4 phase become clearly visible. Obviously the solubility limit for Mn in GaN was exceeded and metal rich phases have been formed. This data also serves to verify what was implied from the growth rate and electrical data. RHEED analysis also showed evidence of a 3D or a polycrystalline structure. However, when GaMnN films with Mn concentrations less than 9 at % were examined, no evidence of any other Mn-containing phase could be found other than GaMnN. Table I shows a compilation of potential second phases that could form in transition metal doped GaN (and other III-V semiconductors) and also their magnetic properties.

Effect of Substrate Type and C Doping on GaMnN Magnetic Properties

It was determined from Figure 1a that a T_Mn of 724 °C should be used to produce a Mn level of approximately 5 at % within the GaMnN. Subsequent experiments were carried out using this Mn flux to try to produce material showing DMS-induced ferromagnetism. In addition, the effect of GaMnN film crystallinity and the effect of carrier concentration on the resulting magnetism were also investigated.

For this portion of the research, (0001) Al₂O₃ and the MOCVD derived GaN would be used as the starting substrate materials. In order to directly compare the results of using more than one type of substrate, and to remove the possibility of the MBE parameters drifting over the course of multiple growth runs, GaMnN was grown on both substrates simultaneously. The MOCVD GaN was n-type to 10¹⁸/cm³, doped with Si donors. The Root Mean Square (RMS) roughness of this surface is approximately 2.2 Å, as determined by Atomic Force Microscopy (AFM). Therefore, any GaMnN grown on top of this surface should be of high crystal quality. Because of the exposed GaN surface, no surface nitridation step or AlN buffer growth was performed prior to the growth of the GaMnN. Due to the GaMnN then being directly grown on the sapphire surface, the approximate 14 % mismatch was assumed to produce a highly polycrystalline film. Again, the films were grown at a temperature of 750 °C with T_Ga = 955 °C and T_Mn = 724 °C, with the same N₂ plasma conditions, for 90 minutes. In the first case the material was grown without adding C. A second identical experiment was performed, this time adding C from a CBr₄ bubbler source. The CBr₄ was operated with a He carrier gas flow of 6.8 sccm and a pressure of 68 torr, while the bubbler recirculating bath was held at a temperature of 0.0 °C. This CBr₄ operating parameter has been previously calibrated by SIMS to produce a C concentration of 10²⁰/cm³ within the GaN. Before the growth was started, the substrates were held at 750 °C for 10 minutes under no nitrogen plasma. This was done as an in-situ clean to remove any oxide from the MOCVD GaN surface. The GaN surface went from a polycrystalline/amorphous RHEED pattern to the streaky 2D RHEED pattern. This assures a high-quality starting surface for growth. Chemical analysis of the material by AES indicates a level of 6.1 at % Mn in the GaMnN sample and 4.5 at % Mn in the GaMnN:C sample. Hall effect measurements on the finished material showed the GaMnN to be highly resistive, while the GaMnN:C was conductive to n » 5 X 10¹⁸/cm³. However, the addition of C typically produces highly resistive material. The fact that n-type GaMnN was observed suggests that the C is behaving as an amphoteric dopant in this material (in the presence of the Mn). The addition of C to GaN has been shown to produce p-type material.^(23,24) An understanding of why the presence of C produces n-type material in the presence of Mn can be found in Figure 3, which shows carrier concentration versus the difference in the III-C and V-C bond energies in a variety of semiconductors. From the Figure, it can be seen that the III-C bond energy is lowered in GaMnN compared to GaN due to the relative weakness of the Mn-C bond (E_Mn-C < E_Ga-C). This will allow it to be energetically favorable for the C to be substitutional on the Ga sublattice, enabling the C to serve as an electron donor.

From x-ray diffraction measurements of the GaMnN and GaMnN:C grown on both sapphire and MOCVD GaN, all of the peaks present in the scans can be attributed to the K_a and K_b peaks of GaN or sapphire. No second phases are present in the material. While no second phases were visible within the material, it becomes important to ascertain what Mn_xN_y phase will appear within the epitaxially grown GaMnN films. In the GaMnN powder material, it was found that Mn₃N₂ was the stable phase at the pressures and temperatures used for growth.^(13,14) A Mn_xN_y x-ray standard was grown at 750 °C using the 350 W/3 sccm N₂ plasma condition used to date for growth. In order to assure a sufficient flux of Mn, T_Mn was 887 °C for this sample growth run. Subsequent x-ray diffraction of this sample is given in Figure 4. Analysis indicates that the thermodynamically stable Mn_xN_y phase at this growth temperature is Mn₄N.⁽²⁵⁾ The K_a and K_b peaks can be identified within this sample to be from the (111) and (222) planes of Mn₄N. X-ray diffraction of the GaMnN has never indicated the presence of these peaks within the limit of detection of this measurement.

GaMnN and GaMnN:C Magnetic Properties

Magnetically, higher carrier concentrations appear to be beneficial, as is prescribed by the bound magnetic polaron theory.⁽⁹⁾ The highly resistive films, despite 6.1 at % Mn, only show paramagnetic behavior. In the films with n = 5 X 10¹⁸/cm³, clear hysteresis in M versus H was observed at a temperature of 10 K, indicative of a ferromagnetic material. This suggests that the Mn-Mn exchange is indeed mediated by charge carriers. The effect of crystal quality on the ferromagnetism of the single-phase films is also apparent. The GaMnN:C film grown directly on sapphire, due to the lack of a buffer and the 14 % mismatch, is assumed to be polycrystalline. This film has a measured saturation magnetization of approximately 0.3 emu/cm³. The same material grown on the MOCVD GaN buffer has a measured saturation magnetization of approximately 2.4 emu/cm³. This suggests that crystallinity plays a significant role in the magnetization since the MOCVD GaN is expected to produce superior growth. This is also evident in the moment of the grown films, ~0.008 Bohr magnetons per Mn for the material on sapphire versus ~0.28 Bohr magnetons per Mn for the material on GaN. This suggests also that a more ordered film leads to enhanced incorporation of Mn ions onto the Ga sublattice. Finally, the polycrystalline versus single crystal nature of the films manifests itself in the values of the coercive field for the two GaMnN:C films. The polycrystalline sample has a higher coercivity (~200 Oe) compared to the single crystal sample (~30 Oe). The higher coercive field suggests that the polycrystalline material is broken up into randomly oriented domains of local ferromagnetic order, compared to the single crystal film, where the carrier mediated exchange can exist over larger distances without being interrupted by grain boundaries. This allows an enhanced collective ferromagnetism, resulting in a lower coercive field. For the polycrystalline sample, the application of a field must align the individual grains, producing more of a magnetic history when the field is reversed.

Gallium Manganese Nitride Growth at Elevated Temperature

From the previous section, the presence of C impurities and their related effect on the carrier concentration of the GaMnN indicates the importance of the carrier concentration on the DMS ferromagnetism. However, the influence of C in the possible formation of a Mn-C complex and any possible related effect on the observed magnetic properties leads to the need to deconvolute C from the GaMnN material. To this end, the GaMnN was grown at an elevated temperature compared to that used in the earlier studies of the GaMnN system. Normally, as-grown GaN is n-type due to the formation of group V (nitrogen) vacancies during epitaxy and also due to oxygen if present in the film. With an elevated growth temperature, this effect will be enhanced, as more of the atomic nitrogen will tend to desorb from the GaMnN surface.

The GaN and GaMnN layers were grown at a substrate temperature of 700,750 or 925 °C, with a K-cell temperature of 620 °C, Ga cell temperature of 865 °C and a GaMnN layer growth rate of 700-1000 Å/sec. The maximum Mn concentration for retaining single phase material increased with decreasing growth temperature-at 700°C the maximum Mn concentration was ~9 at.%, while at 925°C it was about 3 at.%.

Figure 5 shows the resistivity of the GaMnN and GaN films grown under similar conditions, as a function of growth temperature. The resistivity decreases with increasing T_G in both cases. All films were n-type. Smoother films were obtained at higher T_growth, with the smoothest films being the ones grown on MOCVD GaN buffers (Figure 6).

Figure 7 shows the effect of Mn K-cell temperature on GaMnN resistivity and root-mean-square (RMS) roughness. The film roughness increases with increasing Mn flux and at 630°C Ga-Mn phases were detected. The resistivity decreased at high Mn fluxes.

Figure 8 shows the effect of the buffer on the crystallinity of the GaMnN film. Clearly, the use of the MOCVD GaN buffer produces better crystallinity in the overlying GaMnN. Using the optimum growth conditions of T_G =700°C and Mn cell temperature of ~3at.5, no second phases were observed in either high resolution transmission electron microscopy (TEM) or selected area diffraction patterns (SADP), as shown in Figure 9.

Figure 10 shows the result of single axis w-2q rocking curve analysis on films grown on both sapphire substrates and MOCVD GaN buffer layers. As with the q-2q x-ray diffraction and the results from TEM, the w-2q scans do not indicate the presence of second phases within either of the two films. In Figure 10a, the peak at approximately 34.55° corresponds to the (002) reflections from the GaN and GaMnN layers, while the peak at 41.67° corresponds to the (006) reflection of sapphire. The small peak at 35.89° is due to the (002) reflection of the AlN buffer layer. Figure 10b again confirms the q-2q result for the MOCVD sample, prominently showing only the (002) peak from the GaN and GaMnN layers, as well as the (006) peak from the substrate. Due to the large FWHM of the underlying thick (3 µm) MOCVD GaN layer (also seen in as-received GaN), features in 2q due to the GaMnN (002) peak are obscured. The placement of an analyzer crystal in the reflected beam would be required to possibly deconvolute the GaN peak from the GaMnN peak. The absence of an easily discernible knee on either side of the (002) GaN peak in Figure 10a as well indicates a low degree of Mn substitution into the Ga sublattice for this film, despite its observed magnetic properties at low temperatures (see next section). This is confirmed by the calculated moment per Mn ion, which is found to be approximately 0.3 µ_B per Mn, far less than the 5 µ_B per Mn one would expect for 100% Mn substitution.

Electrical and Magnetic Properties

Magnetotransport properties of the GaMnN material were investigated in the temperature range between 10 K and 300 K for magnetic field sweeps between –7 T and +7 T. The Hall (transverse) and sheet (longitudinal) resistances were measured after applying In ohmic contacts to the GaMnN. An AC impedence bridge was used in these measurements to maximize the signal to background noise ratio. The electron carrier density at 300 K was found to be 2.4 X 10¹⁹/cm³, while the carrier density at 10 K was found to decrease slightly to 1.3 X 10¹⁹/cm³. The high carrier density at 300 K indicates that the high growth temperature was effective in increasing the number of free electrons via an excess in nitrogen vacancies, despite the presence of the high concentration of Mn. The sheet resistance showed clear negative magnetoresistance (11% at 10K) as shown in Figure 11. The anomalous Hall effect was observed up to ~20K, which was limited by instrumental resolution. The Hall resistance data was found to be linear from 25 K to 300 K, and is consistent with the absence of magnetic moments due to the thermodynamically stable ferrimagnetic Mn₄N phase, which is reported to have a T_Curie as high as 745 K.⁽²⁶⁾

Figure 12 shows magnetization data obtained from SQUID measurements on films grown at 700°C with different Mn concentrations. There is clear evidence for the presence of magnetization at room temperature, while a Mn concentration of ~3 at.% appears to be optimum at this growth temperature.

SUMMARY AND CONCLUSIONS

We have found that MBE growth of GaN with a high concentration of Mn results in the formation of the ferromagnetic phase GaMnN. The Mn has been found to incorporate linearly with increasing Mn cell temperature for Mn concentrations up to 43 at %. No second phases were found in the GaMnN for Mn concentrations below 9 at %. Both negative magnetoresistance and anomalous Hall effect were observed. The lack of any detected second phases by x-ray diffraction analysis within the GaMnN material in combination with the observed 2D/3D RHEED pattern of the final epitaxial surface indicates that the ferrimagnetic thermodynamically stable Mn₄N phase is not forming within the film. Therefore, we infer that the observed ferromagnetism within the material is due to the formation of GaMnN. This result is significant on several levels. First, ferromagnetism within epitaxial GaMnN is reported. Second, the exchange interaction producing the ferromagnetism in this case is being mediated by electrons and not holes. Although this is possible according to the current theories on DMS ferromagnetism⁽⁹⁾, the predicted T_C’s are generally fractions of a degree K. This raises the possibility of further increasing the T_C within the GaMnN material by refining the growth procedure, while not relying on the presence of holes, which is a problematic issue in III-nitride development.

ACKNOWLEDGEMENTS

The authors would like to thank E. Lambers at the MAIC facility for his assistance with the Auger electron spectroscopy analysis. Support for this work was provided by the U.S. Army Research Office under grant no. ARO- DAAG55-98-1-0216 and by the National Science Foundation under grant nos. DMR-9705224 and DMR-0101438. The work at SNU is partially supported by KOSEF and Samsung Electronics Endowment through CSCMR.

REFERENCES

(1) G. Prinz and K. Hathaway, Phys. Today 48, p. 24 (1995).

(2) D. DiVincenzo, Science 270, p. 255 (1995).

(3) G. Prinz, Science 282, p. 1660 (1998).

(4) D. Deutsch, Proc. R. Soc. Lond. A 400, p. 97 (1985).

(5) B. T. Jonker, Y. D. Park, B. R. Bennett, H. D. Cheong, G. Kioseoglou, and A. Petrou, Phys. Rev. B 62 (12), p. 8180 (2000).

(6) D. D. Awschalom and R. K. Kawakami, Nature 408, p. 923 (2000).

(7) M. Tanaka, J. Crys. Growth 201/202, p. 660 (1999).

(8) B. Lee, T. Jungwirth, and A. H. MacDonald, Phys. Rev B 61 (23), p. 15606 (2000).

(9) T. Dietl, H. Ohno, F. Matsukara, J. Cibert, and D. Ferrand, Science 287, p. 1019 (2000).

(10) H. Ohno, Science 281, p. 951 (1998).

(11) H. Munekata, A. Zaslavsky, P. Fumagalli, and R. J. Gambino, Appl. Phys. Lett. 63 (21), p. 2929 (1993).

(12) F. Matukura, E. Abe, and H. Ohno, J. Appl. Phys. 87 (9), p. 6442 (2000).

(13) M. Zajac, R. Doradzinski, J. Gosk, J. Szczytko, M. Lefeld-Sosnowska, M. Kaminska, A. Twardowski, M. Palczewska, E. Grzanka, and W. Gebicki, Appl. Phys. Lett. 78 (9), p. 1276 (2001).

(14) W. Gebicki, J. Strzeszewski, G. Kamler, T. Szyszko, and S. Podsiadlo, Appl. Phys. Lett 76 (26), p. 3870 (2000).

(15) N. Theodoropoulou, K. P. Lee, M. E. Overberg, S. N. G. Chu, A. F. Hebard, C. R. Abernathy, S. J. Pearton, and R. G. Wilson, J. Nanosci. Nanotech. 1 (1), p. 101 (2001).

(16) M.L.Reed, M.K.Ritums, H.H.Stadelmaier, M.J.Reed, C.A.Parker, S.M.Bedair and N.A.El-Masry, Mater. Lett. 51, 500 (2001).

(17) M.L.Reed, N.A.El-Masry, H.H.Stadelmaier, M.E.Ritums, N.J.Reed, C.A.Parker, J.C.Roberts and S.M.Bedair, Appl. Phys. Lett. 79, 3473 (2001).

(18) N.A.Theodoropoulou, A.F.Hebard, M.E.Overberg, C.R.Abernathy, S.J.Pearton, S.N.G. Chu and R.G.Wilson, Appl. Phys. Lett. 78, 3475 (2001).

(19) S.Sonoda, S. Shimizu, T. Sasaki, Y. Yamamoto and H. Hori, J. Cryst. Growth, 237-239, 1358(2002).

(20) M.E.Overberg, C.R.Abernathy, S.J.Pearton, N.Theodoropoulou, K.T.McCarthy and A.F.Hebard, Appl. Phys. Lett. 79 p. 1312 (2001).

(21) T.Sasaki, S.Sonoda, Y.Yamamoto, K.Suga, S.Shimizu, K.Kindo and H.Hori, J.Appl.Phys.91 p. 7911 (2002).

(22) G.T.Thaler, M.E.Overberg, B.Gila, R.Frazier, C.R.Abernathy, S.J. Pearton, J.S.Lee, S.Y. Lee, Y.D. Park, Z.G. Khim, J. Kim and F.Ren, Appl. Phys. Lett. 80, p. 3964 (2002).

(23) C. R. Abernathy, J. D. MacKenzie, S. J. Pearton, and W. S. Hobson, Appl. Phys. Lett 66 (15), p. 1969 (1995).

(24) W.Schoenfeld, M.J.Antonell and C.R.Abernathy, J.Cryst.Growth 188 50(1998).

(25) R.H.Crabtree, The Organometallic Chemistry of the Transition Metals, 3^rd edition (Wiley, NY, 2001)

(26) A. F. Guillermet, and G. Grimvall, Phys. Rev B 40 (15), p. 10582 (1989).

Table I: Compilation of potential second phases in transition metal doped wide bandgap semiconductors with known magnetic properties.

Phase	Nature of Magnetism	Applicable Magnetic Temperature (K)
Co	Ferromagnetic	1382
Cr	Antiferromagnetic	311
Fe	Ferromagnetic	1040
Ni	Ferromagnetic	627
Mn	Antiferromagnetic	100
Fe3Ga4	Ferromagnetic	483 or 697
Fe2Ga	Ferromagnetic	620
Fe3Ga	Ferromagnetic	760
Fe-Ga alloys	Ferromagnetic
Fe4N	Ferromagnetic	760
FeP	Ferromagnetic	215
Fe2P	Ferromagnetic	278
Fe3P	Ferromagnetic	716
FeP2	Antiferromagnetic	250
Mn2Ga	Ferromagnetic	690
e-Mn3Ga	Ferromagnetic	743
V-Mn5Ga8 (Mn0.6Ga0.4)	Ferromagnetic	210
MnGa	Ferromagnetic	> 300
Mn4N	Ferromagnetic	745
MnP	Ferromagnetic	291
MnP	Antiferromagnetic	50
Mn3P	Antiferromagnetic	115
Mn2P	Antiferromagnetic	103
CrN	Antiferromagnetic	273
Cr2N	Ferromagnetic (?)	Not Ferromagnetic between 85 K and 500 K
Ni3P	Pauli paramagnetic
Ni2P	Pauli paramagnetic
NiP2	Exhibits magnetism
Amorphous Ni-P alloys	Weak homogenous ferromagnetism
CoP2	Diamagnetic semiconductor
CoP	Weak ferromagnetic	<< 1382
V3Ga	Superconductor	Tcritical = 16.8 K

FIGURE CAPTIONS

Figure 1: Variation of Mn concentration (a) and growth rate (b) in GaMnN grown at 750 °C as a function of reciprocal Mn K-cell temperature.

Figure 2: X-ray diffraction of GaMnN with the Mn K-cell at a temperature of (a) 747 °C and (b) 887 °C

Figure 3: Maximum reported carrier concentration versus difference in III-C and V-C bond energy for a variety of compound semiconductors.

Figure 4: X-ray diffraction scan of sample grown at 750 °C with only N plasma and Mn shutters open, identifying the Mn₄N phase. The Mn cell temperature was 887 °C.

Figure 5: Resistivity of GaMnN and GaN as a function of T_G.

Figure 6: RMS roughness (in Å) of GaMnN as a function of T_G.

Figure 7: Resistivity and RMS roughness versus K-cell temperature for a T_G of 700°C.

Figure 8: RHEED patterns from GaMnN films grown on sapphire (top) or MOCVD GaN buffers on sapphire (bottom). The T_growth was 700°C, T_Mn was 620°C and T_Ga in both cases.

Figure 9: High resolution TEM (top) and SADP (bottom) of 3 at.% Mn GaMnN grown at 700°C.

Figure 10: Single axis w-2q rocking curves for GaMnN grown at elevated temperatures on sapphire (a) and MOCVD GaN buffers (b).

Figure 11: Magnetotransport data from GaMnN grown at T_G =925°C on an MOCVD GaN buffer.

Figure 12: M-T (top) and M-H (bottom) from GaMnN films grown at 700°C.The M-T plot is from a sample with ~3 at.% Mn.