↙Chinese
 Home   Services

 
Incorporation of dopants in epitaxial SiC layers grown with fluorinated CVD chemistry
TIME:2017-5-31 15:30:00
Fluorinated chemistry in chemical vapor deposition (CVD) of silicon carbide (SiC) with SiF4 as Si precursor has been shown to fully eliminate the formation of silicon clusters in the gas phase, making SiF4 an interesting Si precursor. However, before a fluorinated CVD chemistry can be adopted, the effect of fluorine on the dopant incorporation must be understood since dopant incorporation is of paramount importance in semiconductor manufacturing. Here, the authors present dopant incorporation studies for n-type doping with N using N2 and p-type doping with Al using TMAl in fluorinated CVD of homoepitaxial SiC. The precursors used were SiF4 as Si precursor and the source of F together with CH4 as C precursor. The authors find that it is possible to control the doping in SiC epitaxial layers when using a fluorinated CVD chemistry for both n- and p-type materials using the C/Si ratio as in standard SiC CVD. However, large area doping uniformity seems to be a challenge for a fluorinated CVD chemistry, most likely due to the very strong Si每F and Al每F bonds.
I. INTRODUCTION
Silicon carbide (SiC) is a wide band-gap semiconductor with outstanding potential for high-temperature, high-power, and high frequency applications.1,2 The hot wall horizontal chemical vapor deposition (CVD) method is the most studied and best suited technique for growing the epitaxial layers which serve as the active region in all electronic SiC devices.3 The demands on the layers differ substantially between different applications, e.g., power devices require thick (30每100 米m), low-doped (1013每1015 cm−3) layers, whereas high-frequency devices require thinner (in the micrometer range) layers with moderate doping (1017 cm−3). The standard approach to SiC CVD is to use silane, SiH4, and small hydrocarbons, e.g., propane, C3H8, or ethylene, C2H4, as precursors diluted in a hydrogen carrier gas flow. The typical growth temperature is 1550每1650 ∼C rendering growth rates in the range 5每20 米m/h range.4 In order to increase the growth rate, mainly for production of high power devices, several routes to high growth rate SiC CVD has been explored. As SiC CVD is done in the mass transport controlled CVD regime, in order to increase the growth rate, the amount of Si and C growth species impinging the substrate surface must be increased. This is typically achieved by increasing the flow of the precursors into the CVD reactor. The main problem with a higher precursor flow is homogeneous gas phase nucleation, mainly the formation of silicon droplets. This can be avoided by performing CVD at a lower total pressure which in turn will yield a lower partial pressure of Si, lowering the probability for Si droplet formation.5 Another route is to add a chemical species that outcompetes the formation of Si每Si bonds by forming bonds to Si stronger than the Si每Si bond. For this route, addition of chlorine has been studied in great detail,6 but it has also been shown that addition of bromine7 or fluorine8 has the same effect.
Of paramount importance when depositing electronic device structures is the ability to control the amount of incorporated dopants in the material. Aluminum and nitrogen are the most common dopants in SiC for p-type and n-type doped layers, respectively. In SiC CVD, these are typically introduced as trimethyl aluminum (TMAl) and dinitrogen (N2). The substitutional incorporation of Al and N in the SiC lattice is controlled by the site-competition theory where Al and N are thought to replace Si and C in the SiC lattice, respectively.9 This makes the C/Si ratio in the CVD gas mixture the key parameter for controlling the amount of incorporated dopants. Doping incorporation in a chlorinated chemistry has been studied in detail, and it was found that n-type doping by N2 and tertbutylphosphine (C4H9PH2) for doping with phosphorus was easily done and not affected by the presence of chlorine,10 while p-type doping by TMAl and triethylboron [B(C2H5)3] were both affected by the presence of chlorine, which limited the maximum net carrier concentration to the low 1018 cm−3 range for Al doping and to the low 1019 cm−3 range for B doping.11 This was suggested to be due to the formation of volatile monochlorides, AlCl and BCl at the CVD conditions, 1570 ∼C and 200 mbar. This is also supported by thermochemical modeling.12 This hypothesis is further supported by a p-type doping study in low temperature CVD of SiC with chlorinated chemistry at 1300 ∼C, where the stability of AlCl is lower,12 where Al doping in the 1020 cm−3 range was measured by SIMS.13 Two overall conclusions from the doping studies with a chlorinated chemistry are that the C/Si ratio can still be used to control the doping and that chlorine seems to decrease the incorporation of p-type dopants. No detailed doping study has been done with a brominated chemistry, but it was reported that the net carrier concentration in the unintentionally doped layers were n-type, where the n-type conductivity stemmed from the background doping in the CVD reactor, decreased with a higher C/Si ratio in the CVD gas mixture into the reactor.7 This behavior is expected from the site competition theory. No doping studies have been reported for a fluorinated chemistry in SiC CVD. In this paper, we present a study on how n-type doping by N2 and p-type doping by TMAl is affected by the presence of fluorine in the CVD process when using SiF4 as Si precursor for SiC CVD.
II. EXPERIMENT
For the growth of epitaxial SiC layers, a horizontal hot wall CVD reactor without rotation was used. The growth chamber was a SiC coated graphite susceptor with a gas cross section of 78 ℅ 24 mm in width and height, respectively, and a length of 190 mm, which was preceded by a SiC coated widened graphite inlet after a gas liner made of quartz. Substrate pieces of 4H-SiC and a 4∼ off-cut angle, 16 ℅ 16 mm in size, were placed at 3, 6, and 9 cm into the susceptor from the susceptor inlet. As precursors SiF4, CH4 and N2 or TMAl were used as Si, C and N or Al precursors, respectively. SiF4 also acted as the source of F. The precursors were diluted in palladium membrane purified H2 acting as the carrier gas. For the experiments, a carrier gas flow of 37.5 standard liters per minute (SLM) at a process pressure of 100 mbar was used. This gives a room temperature corresponding average gas speed velocity of 3.3 m/s in the susceptor. The process temperature was 1600 ∼C and the Si/H2 = 0.125%.
When investigating doping using N2, the C/Si ratio was first fixed at 1.0, and the N2/Si ratio was varied from 0.1 to 10, then the N2/Si was fixed at 1.0, and the C/Si ratio was varied from 0.4 to 1.2 in steps of 0.2. Experiments where the N2/Si was reduced proportional to the C/Si ratio to keep N2/C = 1 as the C/Si was varied from 0.4 to 1.0 were also done. One additional experiment using 5 SLM of N2 was made to supplement the series of increased N2 flow for a N2/Si ratio of 100. To retain the average gas velocity and keeping the process pressure, the amount of carrier gas was reduced from 37.5 to 32.5 SLM, otherwise keeping the flow rates of SiF4 and CH4, thereby resulting in a Si/H2 ratio of 0.144% instead of 0.125%. When investigating doping using TMAl the C/Si ratio was kept at 1.0 while varying the Al/Si ratio in the range from 10−5 to 10−2 and varying in the C/Si ratio in the interval 0.4每1.2 in steps of 0.2 when Al/Si = 10−4.
For the measurement of the net donor and acceptor concentrations, capacitance每voltage measurements were used using a mercury probe station. The thickness of the grown epitaxial layers was approximated from the extension of triangular defects14 on the sample at least 2 mm away from the sample edge to avoid growth rate edge effects. The extension of the triangular defects was measured using optical microscope.

 Contact Us       Links       Copyright explanation   
WWW.CHINESEVACUUM.COM
Addㄩ75-1 Fan Yu Rd.Shanghai China 200052 Email: Webmaster@chinesevacuum.com
Hot: 0086-21-62595067 Tel: 0086-21-62595268 Fax: 0086-21-52906507
QQ Group 1: 286813114, Group 2: 162715660, Group 3: 256587195

Website Chinesevacuum All Rights Reserved.