Transition Elements as Deep Level Dopant

Transition Elements as Deep Level DopantHIGH resistor SILICON DEEP-LEVEL DOPING COMPENSATION USING ELEMENTAL GOLDINTRODUCTION1.1 query BackgroundMo no.ithic nuke integrated circuit (MMIC) is a vaporize circuit in which two active and passive components argon pretended on the same semiconductor substratum 1. The development of MMICs has been augmented by the racy demand for exalted-speed devices operating at microwave relative frequency ranging between 300 MHz and 300 GHz. Their profits of being small, light, and cheap in large quantities return aldepressioned the proliferation of elevated frequency devices such as cellular phones. However, a problem will arise when standard ti (Si) substratum is employ to operate in super utmost frequency surround (SHF). The richly absorption of microwave exponent will be caused by the background free carriers present in the substrate 2. at that placefore, low tone ending and high impedance substrates are chartered to elimi nate the problem. It lav be achieved by reducing the number of background free carriers in the substrate which result in degradation of circuit performance.The III-V semiconductor materials such as GaAs, GaN and InP has been widely used in the take of high impedance substrates referable to their wide exercise setgap nature. However, the wafer diameter begetd employ III-V materials is typically from 4 to 6 3. This increases the bell of production since the standard wafer diameter for modern CMOS engine room is 12 4. Furthermore, the lattice-mismatch problems will complicate the fabrication operation, ca employ the cost to increase. Therefore, Si has been considered to be an alternating(a) material for the III-V semiconductor compound due to less fabrication complexness and cost. However, the background impurities such as vitamin B complex will enter the atomic number 14 during monowatch glassline Si egress causing the increase in substrate losings at microwave range 5.Th ere have been efforts to use the silicon-on-insulator (SOI) technical schoolnologies and silicon-on-anything (SOA) to pound the problem. The SOI wafers seat be produced by several rules silicon-on-sapphire (SOS), separation by imbed oxygen (SIMOX), bond and etch-back SOI (BESOI), Smart Cut 6, and epitaxial socio-economic class conveyancing (ELTRAN) 7. For SOS approach, a thin film of Si is epitaxially grown on sapphire substrate as shown in innovation 1.1. Mean era, the otherwise four approaches use a similar cross section of SOI wafer as shown in form 1.2(b) which consists of lead aims SOI degree (top stage), buried oxide ( boxful) layer (middle layer) and silicon substrate (bottom layer). The purpose of the BOX layer is to electrically insulate a fine layer of SOI layer (where the circuits are placed) from the rest of the Si wafer. The SIMOX approach uses implanted silicon dioxide, SiO2 layer as the BOX layer to separate the top thin Si layer from Si substrate. gens 1 .1 Cross-section of silicon-on-sapphire (SOS) wafer 8 hear 1.2 A conventional representation of bond and etch-back (BESOI) sue 9Apart from the mentioned approaches, BESOI, Smart Cut, and ELTRAN regularitys bes pecker the wafer bonding technique. For BESOI method acting, the thermally oxidised Si wafer (also known as distribute wafer) is bonded to another Si wafer which acts as bond wafer as shown in figure 1.2(a). After the wafer bonding process, the top wafer will be etched to hold in the required thickness for SOI layer as shown in figure 1.2(b). On the other hand, nidation of gas ions, most commonly hydrogen is make after the oxidisation process for Smart Cut method as shown in figure 1.3. The implantation process is meant for layer splitting process to achieve required thickness of SOI layer after the wafer bonding process. The processes twisty in ELTRAN method is shown in figure 1.4. The ELTRAN method uses similar procedures in Smart Cut with the difference in use of double layer porous Si layer instead of implantation of hydrogen ions. The advantage of using Smart Cut and ELTRAN methods is that the initial wafer or seed wafer can be reused for the same process.Figure 1.3 Smart Cut process for production of SOI wafers 9Figure 1.4 ELTRAN process flow based on seed wafer reusage 7Meanwhile, the SOA technology is achieved by gluing a fully-processed SOI substrate to another substrate such as glass and alumina 10. However, the on-chip dissipation which could cause thermal breakdown had been proved to be a severe issue 11. Therefore, there is a need to look for the alternative to SOI and SOA wafers, which is the high ohmic resistance raft Si substrate. In 2009, International Technology Roadmap for Semiconductors (ITRS) had stressed the importance of high ohmic resistance Si in radio frequency (RF) and analog/mixed signal (AMS) CMOS exercise 12.There are generally two techniques for Si crystal growth Czochralski (CZ) technique and float-zone (FZ) technique 13. A simplified version of CZ puller, which is an apparatus used to produce monocrystalline Si ingots for CZ technique is shown in figure 1.5. The high rectitude polysilicon, known as electronic grade silicon (EGS) is melted in a rotating silica or quartz melting pot. A seed crystal is placed in the melt and then slowly withdrawn from the melt. The run silicon adhering to the crystal freezes or solidifies into a continuous crystal from the seed. The diameter of the crystal can be maintained by controlling the temperature of the crucible and the rotating speed of the crucible and the rod. However, the CZ process will come in contamination to the monocrystalline Si due to the presence of oxygen, carbon monoxide and impurities such as type B and phosphorus.Figure 1.5 Czochralski crytal puller. CW represent clockwise rotation and CCW represents counter clockwise rotation 13The FZ process, on the other hand, produces Si crystals with lower contamination as no crucible is used in the process. FZ crystals are mainly used for high power and high voltages devices due to its high immunity. There is a commercially in stock(predicate) high impedance FZ-Si technology called HiRes 14. The pop electrical resistance of Si produced through HiRes is up to 70 k-cm, which is suitable for future GHz and terahertz application. However, it is not suitable for modern CMOS processing since its uttermost wafer diameter is particular to 8, which will increase the cost. Therefore, there is a need to produce high underground bulk CZ-Si substrate due to its low fabrication cost. Therefore, CZ process is still the most widely used method in the manufacturing of ace crystal silicon.In 2003, Mallik et al. 2 introduced a new conceit in underdeveloped a semi-insulating silicon through a method called duncical- take do ensnareg soaking up using 3d conversion elements. It showed that there is possibility to produce high ohmic resistance bulk CZ-Si substrate using occult level doping compensation. Following this produce, Mallik et al. 5 managed to produce CZ-Si bulk substrate using Mn with electric resistance of up to 10 k-cm. Jordan et al. has also used Au to produce CZ-Si wafer with bulk electric resistance of up to 180 k-cm 15. The use of Au-compensated high resistivity bulk Si substrate has been turn out by Nur Z. I. Hashim et al. to be able to seize the parasitic surface conductivity (PSC) effect 16.1.2Problem soilmentThe idea introduced by Mallik et al. 2 on developing high resistivity bulk Si substrate through thick(p)-level doping compensation is solely based on p-type CZ-Si. stock-still though high resistivity bulk Si substrate has been proven to be achievable using p-type CZ-Si, it has been shown in the work by Jordan et al. 15 that high magnitude of Au-compensated high resistivity bulk Si substrate can be achieved by using n-type CZ-Si. The capability and problem of using transformation elements other than Au as the thick(p) level dopants to produce high resistivity bulk n-type CZ-Si substrate have not been discussed by the work mentioned above.1.3Objectives of ResearchThere are three main objectives that must be met in this research fuddleTo investigate the potential of using passing elements as late level dopant for n-type Si substrate as compared to p-type Si substrate.To analyse the result obtained through numerical calculation using MATLAB by comparing it with the data-based data.To make comparative study on the resistivity and effectiveness of the high resistivity bulk substrate produced using n-type CZ-Si with other materials such as III-V semiconductor materials.1.4Scope of ResearchThe scope of this project is to analyse the resistivity plot generated by numerical calculation using MATLAB. The potential and effectiveness of each of the transition elements as dim level dopants for n-type CZ-Si will be discussed in this work. The fabrication and experiment of high resistivity bulk n-ty pe CZ-Si substrate will not be conducted in this work. The experimental data used for comparison with the result of numerical calculation is obtained from Dr. Nur Zatil Ismah Hashim which was do at Southampton Nanofabrication Centre in 2015.LITERATURE REVIEW2.1IntroductionThere are several methods to produce high resistivity bulk Si substrate, namely proton implantation method, helium-3 ion irradiation and cryptical-level doping compensation method. Both proton implantation method and helium-3 ion irradiation use the charge trappings to reduce the true conduction by background free carriers. Meanwhile, the muddy-level doping compensation method use the doubtful dopants to compensate the shoal dopants in the Si substrate.2.2Proton nidation Method one and only(a) of the methods to increase the resistivity of CZ-Si substrate is through proton implantation method. The run of protons into Si bulk structure will create blemishs which can trap wide awake carriers. Therefore, the carrier lifetime is low due to the presence of defects, which prevents the mobile carriers from conducting live freely in the substrate. Table 2.1 summarizes the studies conducted to produce high resistivity bulk CZ-Si substrate using proton implantation method.Table 2.1 Studies on high resistivity bulk CZ-Si substrate using proton implantation method.ContributorYearProton Implantation Energy (MeV)Resistivity of the Produced Substrate (-cm)Li19890.18103Liao et al.199830106Wu et al.200010106Rashid et al.200217.4cvIn 1989, Li 17 managed to produce a high resistivity layer beneath Si surface layer using proton implantation and dance annealing process. The implantation of proton and annealing process formed the buried defect layer with a resistivity of up to 103 -cm. Meanwhile, Liao et al. created semi-insulating regions within silicon substrate with a resistivity of 1 M-cm 18. It was achieved by bombarding proton beams at 30 MeV from a compact ion cyclotron to the surface of Si subst rate. Following this work, the Si substrate with similar resistivity had been produced by Wu et al. using a lower proton implantation strength, which is 10 MeV 19. In 2002, Rashid et al. inform a Si substrate with a resistivity of 0.1 M-cm produced through their six-step implantation method using an implantation energy of 17.4 MeV 20.The high-Q inductors and high transmission gain integrated antenna have been get on the high resistivity Si substrate by Liao et al. and Rashid et al. respectively 18, 20. However, the high process cost is needed for proton implantation method as an enormous dose of 1015 cm-2 to maintain the resistivity of the originally 15 -cm Si substrate to be higher than 1014 -cm 21.2.3Helium-3 Ion Irradiation TechniqueIn 1987, helium-3 ion irradiation technique has been used for carrier lifetime control of silicon power devices 22. The charge trappings created by the helium irradiation and snow scattering of the charged trap will prevent the conduction of curre nt by free mobile carriers in the substrate 23. Therefore, a high resistivity Si bulk substrate can be realised by the reduction in carrier lifetime.In 2014, N. Li et al. reported a high resistivity region created within CZ-Si substrate with a resistivity of over 1.5 k-cm using a dose of 1.51013 cm-2 of helium-3 ions 24. The produced high resistivity Si substrate has been used by N. Li et al. in both work for substrate noise isolation onward motion in a CMOS process and quality factor improvement in on-chip helical inductors 24, 25. R. Wu et al. has also used helium-3 ion irradiation technique in their work on radiation efficiency improvement in 60-GHz on-chip dipole antenna 23.The helium-3 ion irradiation technique has the advantage of saving the product cost up to 97% as compared to proton implantation method by reducing the dose measuring stick from 1.01015 cm-2 to 1.51013 cm-2 24, 25. However, the helium-3 ion irradiation technique is comparably less studied and the problem a ssociated with this technique has not been discussed in the work mentioned above.2.4Deep-Level Doping earnings MethodThe idea of creating high resistivity bulk CZ-Si substrate using deep-level doping compensation has been proposed by Mallik et al. in 2003 2. The basic principle of this method is compensating shallow impurities with deep impurities, i.e. shallow donors are being compensated by deep acceptors (as shown in figure 2.1) whereas shallow acceptors are being compensated by deep donors (as shown in figure 2.2).Figure 2.1 Compensation between shallow donors and deep acceptors 26Figure 2.2 Compensation between shallow acceptors and deep donors 26As illustrated in figure 2.1, deep acceptors introduced an energy level at EA, which is close to the intrinsic Fermi level. The deep acceptors which are negatively charged attract the minority carrier holes to be trapped at EA level. The electrons from shallow donors are initially excited to the conduction luck, then unhorse to EA l evel to recombine with the holes. On the other hand, the positively charged deep donors introduced an energy level at ED as shown in figure 2.2. The minority carrier electrons are attracted and trapped at ED level while the holes from shallow acceptors will fall into valency streak. The trapped electrons at ED level then fall into valency band to recombine with the holes. Therefore, there is no generation of free carriers which reduces the resistivity of the substrate in both cases.Figure 2.3 shows the resistivity of Si at 300K with a background boron tautness of 1014 cm-3, compensated using deep donor impurities with generic wine energy level positions to a lower place conduction band edge, ED. It can be observed that the resistivity of Si increases until a maximal value is reached while the concentration of deep donors, ND increases. The resistivity is low initially due to undercompensation caused by insufficient number of deep donors. The maximum value of the resistivity of Si is reached when deep donors exactly compensate the boron acceptors. Further increase in ND causes overcompensation which results in a fall in the resistivity of Si, making the substrate tends to become n-type.Figure 2.3 mensurable resistivity of Si at 300K as a die hard of generic donor concentration for background boron concentration of 1014 cm-3 2Table 2.2 Positions of energy levels of transition elements in Si 27ElementDonor level below EC (eV)Acceptor level above EV (eV)Co0.890.82Pd0.840.9Au0.780.56Ag0.750.545V0.450.92Mn0.421.0Pt0.3140.889It can be noted that the resistivity peaks are sharper for ED which is lower than 0.3 eV while the resistivity remains high over a range of relatively low concentration determine for larger values of ED. For small values of ED, almost all donors are ionised and take part in the compensation since the donor energy level is nearer to the conduction band than Fermi level. A slight increase in ND causes the resistivity to decrease sharply, c hanging the material to n-type. Meanwhile, for large values of ED, the donor level is near intrinsic Fermi level and less fraction of deep donors is ionised. Therefore, the compensation change gradually with the increase in ND and the resistivity remain high over a wide range of ND.The transition elements are used as deep level dopants as they introduce a pair of deep donor and deep acceptor levels into the Si band gap as shown in table 2.2. The deep dopant energy levels introduced by the transition elements pin the Fermi level near the middle of the Si band gap as shown in figure 2.4 15. Thus, high resistivity CZ-Si substrate can be achieved by capturing the free carriers by deep impurities, which reduces the concentration of background free carrier.Figure 2.4 Fermi level pin by deep levels introduced by transition elements 15Figure 2.5 Calculated resistivity of Si at 300K as a function of Au, Ag, Co and Pd for three different background boron concentration in cm-3 2Figure 2.6 Calc ulated resistivity of Si at 300K as a function of (a) Pt (b) V and (c) Mn concentrations for three different background boron concentrations in cm-3 2Figure 2.7 Calculated resistivity of Si as a function of Au concentration for n-type and p-type Si with a shallow doping concentration of 1014 cm-3 28The transition element dopants are generally grouped into two categories Au, Ag, Co and Pd are in first course of study whereas Pt, V and Mn are in second category. As illustrated in figure 2.5, the resistivity of p-type CZ-Si substrate increases with increasing concentration of the deep dopants in first category. The behaviour of impurities in first category is due to presence of both deep donor and acceptor levels very near intrinsic Fermi level of Si bandgap. For Au and Ag, the resistivity of Si reaches a plateau at the concentration of deep dopants over 1016 cm-3 for three different background boron concentration. The mind for the slight difference in the behaviour of Au and Ag is that the both donor and acceptor level is nearer to the middle in the Si bandgap as compared to Co and Pd.For the second category of deep dopants, the resistivity of p-type CZ-Si reaches a peak and then reduce sharply with the increase in the concentration of the deep dopants as shown in figure 2.6. The reason of the difference in the behaviour is that the impurities in second category have either donor or acceptor level near the intrinsic Fermi level. Therefore, the dopants in second category can only compensate for a single type of drug silicon substrate.The effect of using 3d transition elements as deep level dopants in n-type CZ-Si substrate has not been shown in the work by Mallik et al. Meanwhile, it is shown in the work by Jordan et al. that higher bulk resistivity of Au-compensated Si substrate can be achieved by using n-type CZ-Si as shown in figure 2.7 28. The n-type Au-compensated Si substrate with resistivity up to 70 k-cm has been used by Nur Z. I. Hashim et al. for recognition of coplanar waveguides (CPW) on the substrate 29. Therefore, the potential and problem of using 3d transition elements for deep level compensation in n-type CZ-Si substrate will be discussed in this work.2.5SummaryThe realisation of high resistivity bulk Si substrate using proton implantation method and helium-3 ion irradiation technique was studied. The fabrication of high-Q inductors and antenna has been done on the Si substrate produced using both methods. However, there are problems associated with both methods such as high product cost for proton implantation method and being comparably less studied for helium-3 ion irradiation. Therefore, the idea of creating a semi-insulating silicon substrate using deep-level doping compensation with 3d transition elements was proposed by Mallik et al. in 2003. The deep-level doping compensation method has since been well studied and used for the fabrication of coplanar waveguides and inductors by Nur Z. I. Hashim et al.Reference s1I. D. Robertson and S. Lucyszyn, RFIC and MMIC Design and Technology. IET, 2001.2K. Mallik, R. J. Falster, and P. R. Wilshaw, Semi-insulating silicon using deep level dross doping problems and potential, Semicond. Sci. Technol., vol. 18, no. 6, p. 517, 2003.3Products Capabilities EpiWorks, EpiWorks. Online. on hand(predicate) http//www.epiworks.com/products-capabilities/. Accessed 28-Feb-2017.4Global Manufacturing at Intel, Intel. Online. Available http//www.intel.co.uk/content/www/uk/en/architecture-and-technology/global-manufacturing.html?wapkw=wafer+size_ga=1.16867193.1775779534.1436014173. Accessed 28-Feb-2017.5K. Mallik, C. H. De Groot, P. Ashburn, and P. R. Wilshaw, Enhancement of resistivity of Czochralski silicon by deep level atomic number 25 doping, Appl. Phys. Lett., vol. 89, no. 11, p. 3, 2006.6Smart Cut technology, Smart Choice Soitec, Soitec. Online. Available https//www.soitec.com/en/products/smart-cut. Accessed 28-Feb-2017.7T. Yonehara and K. Sakaguchi, ELTRA N clean SOI Wafer Technology, vol. 4, no. 4, pp. 10-16, 2001.8S. Iwamatsu and M. Ogawa, Silicon-on-sapphire m.o.s.f.e.t.s fabricated by back-surface laser-anneal technology, Electron. Lett., vol. 15, no. 25, pp. 827-828, 1979.9G. K. Celler and S. Cristoloveanu, Frontiers of silicon-on-insulator, J. Appl. Phys., vol. 93, no. 9, pp. 4955-4978, 2003.10R. Dekker, P. G. M. Baltus, and H. G. R. Maas, Substrate transfer for RF technologies, IEEE Trans. Electron Devices, vol. 50, no. 3, pp. 747-757, 2003.11N. Nenadovic, V. DAlessandro, L. K. Nanver, F. Tamigi, N. Rinaldi, and J. W. Slotboom, A back-wafer contacted silicon-on-glass integrated bipolar process. give II. A novel analysis of thermal breakdown, IEEE Trans. Electron Devices, vol. 51, no. 1, pp. 51-62, 2004.12RF and AMS tech for wireless communications, International Technology Roadmap for Semiconductors, 2009. Online. Available http//www.itrs2.net/itrs-reports.html. Accessed 28-Feb-2017.13S. M. Sze, Semiconductor Devices Physic s and Technology, second ed. New York John Wiley and Sons, 2002.14High resistivity silicon for GHz and THz technology, Topsil. Online. Available http//www.topsil.com/en/silicon-products/silicon-wafer-products/hiresTM.aspx. Accessed 28-Feb-2017.15D. M. Jordan, R. H. Haslam, K. Mallik, and P. R. Wilshaw, The Development of Semi-Insulating Silicon Substrates for Microwave Devices, Electrochem. Soc., vol. 16, no. 6, pp. 41-56, 2008.16N. Z. I. Hashim, A. Abuelgasim, and C. H. De Groot, Suppression of parasitic surface conduction in Au-compensated high resistivity silicon for 40-GHz RF-MMIC application, 2014 Asia-Pacific Microw. Conf., pp. 55-57, 2014.17J. Li, myth semiconductor substrate formed by hydrogen ion implantation into silicon, Appl. Phys. Lett., vol. 55, no. 21, pp. 2223-2224, 1989.18C. Liao et al., Method of creating local semi-insulating regions on silicon wafers for device isolation and fruition of high-Q inductors, IEEE Electron Device Lett., vol. 19, no. 12, pp. 461-462, 1998.19Y. H. Wu et al., Fabrication of very high resistivity Si with low loss and cross talk, IEEE Electron Device Lett., vol. 21, no. 9, pp. 442-444, 2000.20A. B. M. H. Rashid, S. Watanabe, and T. Kikkawa, High transmission gain integrated antenna on highly high resistivity Si for ULSI wireless interconnect, IEEE Electron Device Lett., vol. 23, no. 12, pp. 731-733, 2002.21L. S. Lee et al., Isolation on Si wafers by MeV proton bombardment for RF integrated circuits, IEEE Trans. Electron Devices, vol. 48, no. 5, pp. 928-935, 2001.22W. Wondrak and A. Boos, Helium Implantation for Lifetime function in Silicon Power Devices, ESSDERC 87 17th Eur. square(a) State Device Res. Conf., pp. 649-652, 1987.23R. Wu et al., A 60-GHz efficiency-enhanced on-chip dipole antenna using helium-3 ion implantation process, 2014 forty-fourth Eur. Microw. Conf., pp. 108-111, 2014.24N. Li et al., High-Q inductors on locally semi-insulated Si substrate by helium-3 bombardment for RF CMOS integrated circuit s, 2014 Symp. VLSI Technol. Dig. Tech. Pap., pp. 1-2, 2014.25N. Li et al., Substrate noise isolation improvement by helium-3 ion irradiation technique in a triple-well CMOS process, 2015 45th Eur. immobile State Device Res. Conf., pp. 254-257, 2015.26J. D. Puksec and V. Gradisnik, Influence of shallow impurity on steady-state chance function of multilevel deep impurity, 2000 10th Mediterr. Electrotech. Conf. Inf. Technol. Electrotechnol. Mediterr. Countries. Proc., vol. I, pp. 185-188, 2000.27W. Schroeter and M. Seibt, Deep Levels of Transition metal Impurities in c-Si, in Properties of Crystalline Silicon, R. Hull, Ed. London INSPEC IEE, 1999, p. 561.28D. M. Jordan, K. Mallik, R. J. Falster, and P. R. Wilshaw, Semi-Insulating Silicon for Microwave Devices, Solid State Phenom., vol. 156-158, pp. 101-106, 2009.29N. Z. I. Hashim, A. Abuelgasim, and C. H. De Groot, Coplanar waveguides on gold-doped high resistivity silicon for 67-GHz microwave application, RFM 2013 2013 IEEE Int. R F Microw. Conf. Proc., pp. 274-277, 2013.