Replacing silicon dioxide as a gate dielectric material in present day complementary metal-oxide-semiconductor technology is an active area of research. This is because fundamental concerns have emerged regarding the increased leakage current and reduced dielectric reliability as a result of rapidly...
Replacing silicon dioxide as a gate dielectric material in present day complementary metal-oxide-semiconductor technology is an active area of research. This is because fundamental concerns have emerged regarding the increased leakage current and reduced dielectric reliability as a result of rapidly shrinking SiO_(2)-based gate dielectric dimensions. This situation has led to the search for an alternative material with a higher dielectric constant than SiO_(2), but with a suitably large band gap like SiO_(2), to keep the gate leakage current within reasonable limits. Amorphous Al_(2)O_(3), which has a dielectric constant of about 8-10 and a band gap of about 7-9 eV, is an attractive alternative for a gate oxide material. To determine the usefulness of Al_(2)O_(3) film in future CMOS devices, it is important that the film be thermally stable on the Si surface. However, it is known that depositing an Al_(2)O_(3) film on the Si surface often results in the formation of an interfacial aluminum silicate and/or a SiO_(2) layer. The Al_(2)O_(3)-Si interface is expected to be stable against the formation of silicate and/or SiO_(2), as such interfacial reactions are not predicted from bulk equilibrium thermodynamics of the Si and Al_(2)O_(3). Bulk thermodynamics is not always a good predictor of interfacial reactions on an atomic scale. Therefore, to obtain a s1;able Al_(2)O_(3)-Si interface, it is important to understand the kinetics of deposition and interface formation. We grew thin Al_(2)O_(3) films on Si (100) surface using atomic layer deposition (ALD) and plasma enhanced ALD (PEALD), and investigated the Al_(2)O_(3)-Si interface structure and stability for each case. In ALD process, we used tri-methyl-aluminum (TMA) or methyl-pyrrolidine-alane (MPA) as A1 sources and used iso-propyl-alcohol (IPA) or H_(2)O vapor a s oxygen sources. In PEALD process, Al_(2)O_(3) films were grown by using MPA and O_(2) plasma, The difference in saturated deposition thickness/cycle of each ALD process was explained by the adsorption of Al sources and secondary adsorption of oxygen sources. Because the effective area of one adsorbed molecule is smaller in the case of MPA than of TMA, the amount of adsorbed MPA in unit area would be more than that of adsorbed TMA. Therefore the saturated deposition thickness/cycle would be increased when MPA was used a s Al precursor. When oxygen source was supplied to the reactor, the IPA or H_(2)O would react with Al precursor previously adsorbed on the surface and form Al_(2)O_(3) The oxygen sources are then likely adsorbed on both of the outermost surfaces of the newly formed Al_(2)O_(3) and the newly exposed surface, which had been screened with adsorbed Al precursor before. This is referred to a s the secondary adsorption. These secondary adsorbed oxygen sources would also react with Al precursor during the following Al precursor supply. H_(2)O is more effective in the secondary adsorption the saturated deposition thickness/cycle would be increased when H_(2)O is used a s oxygen source. The stoichiometry and the carbon contamination of as-deposited and annealed Al_(2)O_(3) films were characterized by X-ray photoelectron spectroscopy (XPS) and Auger electron spectrometry (AES). The interface stability of the Al_(2)O_(3)-Si systems was characterized by cross-sectional transmission electron microscopy (XTEM) and XPS. The formation of interfacial layer in as-deposited films depended upon the oxygen source used in process. An interfacial layer was not detectable on the as-deposited film grown by TMA-IPA or MPA-IPA ALD process. On the other hand, an interfacial layer with a thickness of - 12 Å was generated on the as-deposited film that was grown by ALD using H_(2)O instead of IPA. Similarly, an interfacial layer was formed on the as- deposited film that was grown by PEALD using O_(2) plasma a s oxygen source. H_(2)O vapor or O_(2) plasma oxidized Si surface at a deposition temperature until the surface was completely covered with deposited Al_(2)O_(3) layer. After annealing at 800℃ for 5 min, an interface layer was newly formed or grown even under the neutral ambient of Ar, and it grew thicker under the oxidizing ambient of O_(2). Oxygen, which is needed for the formation of the interface layer during the annealing process, was supplied from bath of the ambient oxygen and the excess oxygen in the films. We proposed the A1N interlayer to protect Si surface from oxygen source during an early stage of Al_(2)O_(3) ALD/PEALD and excess oxygen in the films during post-deposition annealing. The AIN films were deposited by PEALD using MPA and NH_(3) plasma. XTEM results showed the as-deposited and annealed Al_(2)O_(3)-Si interface was abrupt with the introduction of ultra-thin AIN interlayer (∼ 10 Å) before Al_(2)O_(3) PEALD. The effective dielectric constant was improved from 8 of the film without AIN interlayer to 9.6 of the film with AIN interlayer.
Replacing silicon dioxide as a gate dielectric material in present day complementary metal-oxide-semiconductor technology is an active area of research. This is because fundamental concerns have emerged regarding the increased leakage current and reduced dielectric reliability as a result of rapidly shrinking SiO_(2)-based gate dielectric dimensions. This situation has led to the search for an alternative material with a higher dielectric constant than SiO_(2), but with a suitably large band gap like SiO_(2), to keep the gate leakage current within reasonable limits. Amorphous Al_(2)O_(3), which has a dielectric constant of about 8-10 and a band gap of about 7-9 eV, is an attractive alternative for a gate oxide material. To determine the usefulness of Al_(2)O_(3) film in future CMOS devices, it is important that the film be thermally stable on the Si surface. However, it is known that depositing an Al_(2)O_(3) film on the Si surface often results in the formation of an interfacial aluminum silicate and/or a SiO_(2) layer. The Al_(2)O_(3)-Si interface is expected to be stable against the formation of silicate and/or SiO_(2), as such interfacial reactions are not predicted from bulk equilibrium thermodynamics of the Si and Al_(2)O_(3). Bulk thermodynamics is not always a good predictor of interfacial reactions on an atomic scale. Therefore, to obtain a s1;able Al_(2)O_(3)-Si interface, it is important to understand the kinetics of deposition and interface formation. We grew thin Al_(2)O_(3) films on Si (100) surface using atomic layer deposition (ALD) and plasma enhanced ALD (PEALD), and investigated the Al_(2)O_(3)-Si interface structure and stability for each case. In ALD process, we used tri-methyl-aluminum (TMA) or methyl-pyrrolidine-alane (MPA) as A1 sources and used iso-propyl-alcohol (IPA) or H_(2)O vapor a s oxygen sources. In PEALD process, Al_(2)O_(3) films were grown by using MPA and O_(2) plasma, The difference in saturated deposition thickness/cycle of each ALD process was explained by the adsorption of Al sources and secondary adsorption of oxygen sources. Because the effective area of one adsorbed molecule is smaller in the case of MPA than of TMA, the amount of adsorbed MPA in unit area would be more than that of adsorbed TMA. Therefore the saturated deposition thickness/cycle would be increased when MPA was used a s Al precursor. When oxygen source was supplied to the reactor, the IPA or H_(2)O would react with Al precursor previously adsorbed on the surface and form Al_(2)O_(3) The oxygen sources are then likely adsorbed on both of the outermost surfaces of the newly formed Al_(2)O_(3) and the newly exposed surface, which had been screened with adsorbed Al precursor before. This is referred to a s the secondary adsorption. These secondary adsorbed oxygen sources would also react with Al precursor during the following Al precursor supply. H_(2)O is more effective in the secondary adsorption the saturated deposition thickness/cycle would be increased when H_(2)O is used a s oxygen source. The stoichiometry and the carbon contamination of as-deposited and annealed Al_(2)O_(3) films were characterized by X-ray photoelectron spectroscopy (XPS) and Auger electron spectrometry (AES). The interface stability of the Al_(2)O_(3)-Si systems was characterized by cross-sectional transmission electron microscopy (XTEM) and XPS. The formation of interfacial layer in as-deposited films depended upon the oxygen source used in process. An interfacial layer was not detectable on the as-deposited film grown by TMA-IPA or MPA-IPA ALD process. On the other hand, an interfacial layer with a thickness of - 12 Å was generated on the as-deposited film that was grown by ALD using H_(2)O instead of IPA. Similarly, an interfacial layer was formed on the as- deposited film that was grown by PEALD using O_(2) plasma a s oxygen source. H_(2)O vapor or O_(2) plasma oxidized Si surface at a deposition temperature until the surface was completely covered with deposited Al_(2)O_(3) layer. After annealing at 800℃ for 5 min, an interface layer was newly formed or grown even under the neutral ambient of Ar, and it grew thicker under the oxidizing ambient of O_(2). Oxygen, which is needed for the formation of the interface layer during the annealing process, was supplied from bath of the ambient oxygen and the excess oxygen in the films. We proposed the A1N interlayer to protect Si surface from oxygen source during an early stage of Al_(2)O_(3) ALD/PEALD and excess oxygen in the films during post-deposition annealing. The AIN films were deposited by PEALD using MPA and NH_(3) plasma. XTEM results showed the as-deposited and annealed Al_(2)O_(3)-Si interface was abrupt with the introduction of ultra-thin AIN interlayer (∼ 10 Å) before Al_(2)O_(3) PEALD. The effective dielectric constant was improved from 8 of the film without AIN interlayer to 9.6 of the film with AIN interlayer.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.