Plasma deposited diamondlike carbon on GaAs and InP

https://ntrs.nasa.gov/search.jsp?R=19850002536 2020-03-20T21:19:06+00:00Z Metadata, citation and similar papers at core.ac.uk CORE Provided by NASA Technical Reports Server General Disclaimer One or more of the Following Statements may affect this Document This document has been reproduced from the best copy furnished by the organizational source. It is being released in the interest of making available as much information as possible. This document may contain data, which exceeds the sheet parameters. It was furnished in this condition by the organizational source and is the best copy available. This document may contain tone-on-tone or color graphs, charts and/or pictures, which have been reproduced in black and white. This document is paginated as submitted by the original source. Portions of this document are not fully legible due to the historical nature of some of the material. However, it is the best reproduction available from the original submission. Produced by the NASA Center for Aerospace Information (CASI) L. J f • f ,^ NASA Technical Memorandum 86870 FLASISA DEECSIZEU (NASA--rm-86870) DiatlUNDLIKL CANEGN CN GdAs ANC InP (NASA) CSCL 15 p HC AU2/hf A01 N85 -1C844 20i. Uncld s ' G3/76 24199 Plasma Deposited "Diamondlike" Carbon on GaAs and InP J. D. Warner, J. J. Pouch, S. A. Alterovitz, and D. C. Liu Lewis Research Center Cleveland, Ohio and W. A. Lanford SONY-Albany Albany, New York ti ;;^,v 00 104^^ r Prepared for the Thirty-first National Symposium of the American Vacuum Society Reno, Nevada, December 3-7, 1984 LN t .1 r I. INTRODUCTION There has been considerable interest in the properties of carbon f prepared by ion-beam sputtering gases. 3-5 1-6 and rf plasma decom;:sition of hydrocarbon Information on film hardness and bonding arrangements (relative to diamond) can be found in the literature. 2-6 Recently, some optical and electrical characteristics of carbon films grown on Si, which may be useful for semiconductor applications fabrication, have been reported. 1,4,8,9 We have utilized a rf plasma formed from methane gas to deposit the carbon films. The physical and chemical properties of these films have been determined 2 using AES, XPS, ellipsometry, nuclear reaction techniques, IR and UV-vis-NIR i spectrometries. Our results provide information on the growth parameters needed to achieve carbon films with a particular bandgap and hydrogen content., In addition, we demonstrate the importance of the growth temperature on the deposition rate and composition. II. EXPERIMENT a. Growth and Sample Preparation Carbon films were prepared by subjecting the substrates to a rf plasma discharge generated at 30 kHz. We used a Technics PD-II deposition system. All samples were first cleaned in acetone and ethanol baths, and rinsed in deionized water. The samples were then placed in the rf chamber; the background pressure was typically 20 mtorr. Methane gas (99.97 percent pure) was used to flush the chamber several times prior to each run. The power and flow rate settings were varied from 30 to 240 W and from 30 to 90 sccm, respectively; substrate temperatures ranged from ambient (-23° C) to 250° C. We noted that a minimum pressure was required to maintain a stable plasma and it scaled directly with rf power. i s •. Z We report the measured quantities in flow rates with units of standard cubic centimeters (sccm) rather than pressure since it was the flow rate which was controlled during deposition. It was found that the pressure varies S' Linearly with flow rate (f), in the region 30 to 90 sccm. To convert flow rate from sccm to pressure in mtorr we used the formula (1) P - 3.5f + 70 hk where P is the pressure in mtorr. Carbon films were grown on GaAs, InP, Si wafers, quartz, and Multiple Internal Reflection (MIR) plates of Ge and Si. All samples were stored either under vacuum or in a dry box. b. AES/ESCA The composition analyses experiments were performed using a PHI AES/ESCA a system, interfaced with a MACS (Multiple-Technique Analytical Computer System). The sample chamber was evacuated to about 10 H -10 torr prior to the Auger electron profiling and X-ray photoemission measurements. The sputter profiling was done with 2 or 3 KeV Ar ions at 25 mA. The XPS measurements were carried out using a 10 KeV Al Ka x-ray source. c. IR Determination of the relative amount of hydrogen in the film was performed using a ratio recording infrared spectrometer with a data station for data acquisition and analysis of the peak area. The area under the absorption peaks at 2921, 1444, and 1374 cm -1 were used. These measurements were taken in the spectrometer using Ge and Si MIR plates with a 30 0 bevel on the two ends. In order to reduce the edge effects and assure the uniformity of the film across the MIR plates, an aluminum plate of approximately the same thickness as the MIR plates was placed around the plates during deposition. These peaks gave the main vibrational and bending modes of the C-H x (x = 1 to 3) bondings. 2 tin d. i'~ z Optical Absorption Optical absorption measurea^ents were done on carbon films deposited on fused quartz plates to determine the bandgap. The films were measured in a UV-vis-NIR spectrometer with the sample beam perpendicular to the film. The wavelength was scanned from 185 to 836 nm with a bandwidth of 1 nm and a scan rate of 5 nm/sec while the background was automatically subtrated. e. Ellipsometry Optical properties such as the refractive index and the thickness of the r films were determined using a semi-automatic Gaertner L119X research ellipsometer in a rotating analyzer mode. Data acquisition and analysis were done using a computer. At each wavelength, up to nine angles of incidence were used, as needed. The light sources were a He-Ne laser and a 100 W Hg arc source. The photomultiplier readings were averaged over at least 50 rotations, with a measurement every 5 6 of the analyzer, at each wavelength and angle of incidence. The thickness and the complex refractive index at 632.8 nm were measured systematically on a large number of samples. On each sample, data was obtained at angles of incidence in the range of 60 0 to 78 0 . From these measurements * and a were calculated by Fourier analysis. In some samples several wavelengths were used. Then the refractive index and the thickness were calculated from a multiple angle and wavelength (MAW) program developed at the University of Nebraska. substrate were taken from literature. 11 10 The optical data for the A three-phase model was used, i.e., ambient, film, and substrate. This model was found very adequate for our films. f ^ ': c { f. Nuclear Reaction Determination of the absolute amount of hydrogen was performed at the Suny-Albany nuclear accelerator. This method makes use of the 4.43 NeV Y-rays emitted from the H I (N is ,aY)C 12 . reaction. This reaction has an appreciable cross section for N 15 nucleon only ions at 6.385 NeV. The method of calculating the hydrogen concentration from this reaction is given in Ref. 12. III. RESULTS AES and XPS measurements indicated that the films contained only carbon; no other element was observed to the detection limits (0.1 at. ) of the instrument. Figure 1 shows a typical AES profile of carbon films on InP and GaAs. Oxygen was present only on the surface of the films for the GaAs samples whereas for the InP there was a few percent of oxygen at the carbon-InP interface. This suggests that the methane plasma removes all of the oxygen from the GaAs surfaces and most of it from InP surfaces. Our analysis of the percent diamondlikeness of our films comes from the measurement of the asymmetry in the carbon is XPS lineshape due to delocalized ,r-bonds. 13 To quantify this measurement, we compare the relative asymmetry in the C is spectrum of our samples to that of Highly Oriented Pyrolytic Graphite ('HOPG). We define the percent diamondlikeness (DL) to be DL = (1 - al /a2 ) x 100 where al is the area of the tail relative to the total area of the symmetric peak for the sample, and a 2 is the area of the tail relative to the total area of the symmetric peak for HOPG. The measurement value for a2 is 0.25. In Figure 2, we show a comparison of the XPS peaks of HOPG, a a (2) plasma deposited carbon film, and a Gaussian curve fitted to the low energy side of the peak for the carbon film. The DL calculated for the various carbon films ranged from 40 to 97 percent. There is a general trend of decreasing DL values with increasing power/flow rate values. The XPS evaluation of the bonding of the carbon to GaAs and InP were performed with films approximately 200 A thick. The results for a typical film on InP is shown in Fig. 3. The interface was extending approximately from 5 to 20 min ESCA profiling time. Data were taken at 8 and 14 min. All XPS peaks are broadened and shifted in binding energy for material at the interface, as compared to the bulk. This includes the C Is, the In 3d and P 2p peaks, indicating some bonding between the carbon and the InP. Similar behavior was found at the carbon-GaAs interface. Using a rf power of 100 N and a flow rate of 50 sccm with Si and Ge substrates it was found that the carbon growth rate and hydrogen concentration decreases with increase in substrate temperature as measured by the N15 nuclear reaction. Figure 4 shows for carbon on Si that the deposition rate falls off exponentially with temperature whereas the hydrogen concentration decreases linearly with temperature. Deposition rates were determined from measurements of film thickness by a standard profilometry technique. For InP and GaAs, it was found that the carbon growth rate decreases with increasing temperature and above 200° C nucleation does not occur. The carbon films on Si and Ge MIR plates show a decrease in the intensity of the infrared absorptive peaks, with increasing temperature. The IR data show peaks around 2921 cm -I for C-H x (x - 1 to 3) stretzhing modes, peaks at 1445 cm-1 and 1370 cm-1 for the C-H 2 and C-H3 bending modes, and a broad peak at 1610 cm- 1 which is probably associated with amorphous carbon 5 nodes. The absorption intensities of the C-H x peaks can be used to determine the hydrogen concentration once the thickness and optical absorption cross section of the C-H x peaks are known. The optical absorption cross section for these films will be reported elsewhere. 14 Figure 5 illustrates the sensitivity of the IR optical absorption intensity of the C-H x peaks due to changes in Ge substrate temperature. The ellipsometric measurements of index of refraction (n) and extinction coefficient (k), for films produced at rf powers of 50 and 240 W and methane flow rates of 30 to 90 sccm, together with deposition rates from the ellipsometric measured thickness, are given in Table I. The growth rate increased for both rf power and methane flow rate. The index of refraction decreased slightly with flow rate and increased with power. The extinction coefficient at 6328 A was nonzero only at the lowest flow rate whereas at shorter wavelengths for the 50 W and 90 sccm samples, k was measurable but is still gvite small. The change in the index of refraction in samples prepared at different power levels could be due to changes in concentration of voids from sample to sample. To check this we used the effective medium approximation 15 to calculate the concentration of the voids and the required AES profiling time for several samples. Assuming that the 240 W, 50 sccm sample has no voids we calculate the percent voids (PV) in other samples. For example, we found PV = 45 for the 50 W, 90 sccm sample. This value leads to calculated AES profiling time of 108 min as compared with the experimental value 105 min. We thus obtained a ratio of 3.5 for the profiling time, as compared to a factor of 2..0 for the thickness ratio. The optical bandgap results for various power flow rate ratios are given in figure 6. They range from 2.07 to 2.38 eV and decrease with increase in 6 ice^;.3. _ ,F^ .^ ^^ . I- I r power/flow rate for a given power. This tendency is the same as for the percentage diamondlike behavior shorn in figure 3. IV. SUMMARY We have demonstrated that reasonably high purity and quality carbon film can be produced on GaAs and InP over a wide range of rf power and methane flow rate. These films show no impurities other than hydrogen. We found a tendency for the carbon films to be more diamond at lower ratios of rf power to methane flow rates. We were able to control the bandgap and growth rate of the films through reproducible settings of power and flow rate. We have shown the exponential decrease of the growth rate and a decrease of the hydrogen 1 e concentration with increase in substrate temperature during deposition on Si and Ge substrates. Also, we observe a critical temperature (200' Q above i which no nucleation of carbon occurs on GaAs and InP. t More work is needed on carbon films prepared by rf glow discharge to fully determine the variation of the bandgap and growth rate during deposition with the parameters rf power, methane flow rate, and temperature In addition, a the relationship between the bandgap, index of refraction, and the amount of either voids or hydrogen in the films should be investigated. ACKNOWLEDGMENTS We would like to thank John A. Woollam and collaborators at University Nebraska-Lincoln for fruitful discussions and technical help. Y 1 ^] iE^ iMx Yi^ ^. ^^ ^ r *^ .^IT:.^1^Y'. wy} jaifip s REFERENCES 1. A. A. Khan, J. A. Woollam, Y. Chung, and B. Banks, IEEE Electron Device Letters, 4, 146 (1983). 2. C. Weissmantel, K. Bewilogua, D. Dietrick, H. J. Erler, H. J. Hinneberg, S. Klose, W. Nowick and G. Reisse, Thin Solid Films, 72, 19-31, (1980). 3. C. W. Weissmantel, G. Reisse, H. J. Erler, F. Henny, K. Bewilogua, U. Ebersback, and C. Shurer, Thin Solid Films, 63, 315-325 (1979). 4. D. Mathine, R. 0. Dillon, A. A. Khan, G. Bu-Abbud, J. A. Woollam, D. C. Liu, B. Banks, and S. Domitz, J. Vac. Sci. Technol., A2, 365-366 (1984). S. T. J. Moravec and T. W. Orent, J. Vac. Sci. Technol. 18, 226-228 (1981). 6. C. Weismantel, K. Bewilogua, K. Brewer, D. Dietrick, U. Ebersback, H. J. Erler, B. Rau, and G. Reisse, Thin Solid Films, 96, 31-44 (1982). 7. S. Aisenberg and R. Chabot, J. Appl. Phys., 42, 2953-2958 (1971). t. 8. R. 0. Dillon, J. A. Woollam, and V. Katkanant, Phys. Rev. B29, 3482-3489 (1984). 9. Y. Ichinose and F. Shimokawa, in Extended Abstracts and Program. 16th Biennial Conference on Carbon, (American Carbon. Society, University Park, Penn, 1983). 10. G. H. Bu-Abbud, S. A. Alterovitz, N. M. Bashara, and J. A. Woollam, J. w; Vac. Sci. Technol. Al, 619-620 (1983). 11. D. E. Aspnes, and A. A. Studna, Phys. Rev. B27, 985-1009 (1983). 12. T. T. P. Cheung, J. Appl. Phys., 53, 6857-6862 (1982). 13. W. A. Lanford, and M. J. Rand, J. Appl. Phys. 49, 2473-2477 (1978). 14. J. D. Warner, W. A. Lanford, J. J. Pouch, and S. A. Alterovitz, to be i published. 15. D. E. Aspnes, Thin Solid Films, 89, 249-262 (1982). 8 ORIGINAL OF POOR QUALITY TABLE I. - ELLIPSOMETRY DATA OF CARBON FILMS AT 6328 A Sample Power, M Flow rate. cm SO 50 30 50 70 90 50 70 90 50 50 240 240 240 Deposition rate, A/min 79.5 83.3 96.7 100.0 197 263 308 Index of refraction. n Extinction coefficient, k 1.58±0.02 1.71±.02 1.67±.01 1.59±.02 2.15±.04 2.08t.09 1.811.01 -0.01 0 0 0 -.015 0 0 s! 1 a 1r' li *f4 ENAL OFORIGP002 A In 3d P 6 z C Is 4 A z 2 /-0 15 0 - 0 S 16 8 '0P 24 32 40 SPUTTERING TIME, min 4 la C Is C 2 01 lb ), olss 1 r-a 33d Ga 3 A 3d As Ols 0 30 20 10 40 50 SPUTTERING TIME, min Figure 1. - Auger electron spectroscopy IAES) profiles of carbon films on InP (curve A) and GaAs icurve B). using 25 mA. 3 keV Ar ions. 8 fi g ri 6 4C gg f. 2 jA B C C1 0 M -290 -298 -286 I -284 -282 -280 BINDING ENERGY, eV Figure 2. - X-ray photoemission spectroscopy (XPS) of the Intensity of photoemission electrons versus binding energy. (A) - highly oriented pyrolitic graphite 1B1- carbon film on InP, (C^ - fitted gausslon curve to the lower energy side of curve IBL V, ORIGINAL PAGE IF, OF POOR QUALITY 8 PURE CARBON ;i 6 z m AFTER 14 min OF PROFILING i; 4 .o W. S _W Z 2 ' — AFTER 8 min OF PROFILING NO- I 1 I I -296 -292 -288 -284 -280 BINDING ENERGY, eV (a) C is peaks, 6 AFTER 8 min OF PROFILING z > 4 3 /-AFTER FI min OF PROFILING m m W 2 Z InP SUBSTRATE :-, 1 1 458 -454 -450 -446 -442 -438 BINDING ENERGY, eV (b) In 3d peaks. 8 AFTER 8 min r AFTER 14 min OF PROFILING i OF PROFILING 4 InP SUBSTRATE- / W ^. z -144 -140 -136 -132 -128 -124 BINDING ENERGY, eV k (c) P 2p peaks. Figure 3. - XPS intensity versus binding energy at different depth of profiling time for carbon film of thickness of 200 A on InP. 3 i i k 1 .,/ 77'17 11 . r , ORIGIN AL Of POOR QUALI fi i 9x1022 z 0 E k 8 .3 1.5 B 1.0 7 --^ 0 ^ 5 0 ¢ q 1 lu J 1 6 5 50 0 100 150 200 250 300 TEMPERATURE, °C Figurek - Logarithmic plot of deposition rate versus temper ure (A) and hydrogen concentration pH (atoms per cml versus temperature 10) for carbon on a SI substrate with growth parameters: 100 W. 50 sccm, and 30 min. it DEPOSITION TEMPERATURE, oC N H Z } d' C CD C S i-H Z O H d O N m < C 3200 3000 2800 2600 WAVENUMSER, cm-1 Figure 5. - IR absorption Intensity versus wavenumber for carbon on Ge MIR (multiply internal reflection) plates for 7 deposition temperatures at 100 W, 50 sccm, and 30 min. y k i 2 } j i ^I ORIGINAL PAGE M1 OF POOR QUALI 'F'Y a 24 2, 3 d a 0 z 2.2 a m z zo I I I I 3 4 5 RATIO OF rf POWER TO FLOW RATE, Wlsccm i 6 Figure 6. - Optical bandgap versus the ratio of rf power (W) and flow rate (sccm) for different power levels. L