Next generation semiconductor device technology node logic and advanced memory fabrication etch [1] deposition [2] and functionalization [3,4] unit processes require management of process chamber species’ nature, flux and energy at all facets of a substrate and feature being fabricated. Management of profile (conformality), selectivity, deposition or etch rate, uniformity, damage and/or sub-surface and film “quality” needs via continuous or quasi-continuous processes is difficult as process chemistries and plasma source characteristics drive trade-offs innate to their combination. Fortified by more than two decades of progress in modelling and simulation, concurrent process and equipment development has overcome these difficulties. It is evident now to meet trade-off born difficulties in the future that self-limiting process technology will need to be used. In this case, chemistry control may mean parsing individual process steps into adsorption and desorption components [5]. Segregating processes include rapid cyclic etch and atomic layer deposition (ALD). Other examples of self-limiting processes include radical driven plasma doping. This presentation will discuss each of these processes in the context of their needs and how simulation drives their development. Simulation models discussed are based on an integrated approach (chamber through feature) with fluxes to surfaces from plasma source simulations (e.g., HPEM [6]) mated to feature scale simulations such as MCFPM. [7] Atomic layer etching for front end applications will be described that use this suite of simulation models. [8] While atomic layer deposition is more mature than it’s etch counterpart, plasma assisted ALD comes with mechanistic complexity both in the plasma and at surface. [9] Fortunately, quantum chemistry methods, [10] surface and plasma chemistry models can be linked to equipment simulations for concurrent process design. Three dimensional devices require source drain engineering to being about conformal electrical properties. An example is fin extension doping. Radical driven plasma doping can be conformal because of the self-passivating nature of a saturation layer created all around the surface of the fin during the doping process. Functionalizing topographic structures such as silicon, Ge or III-V fins sets severe constraints on the plasma itself. Fins must not erode but sufficient ion energy is required to disorder a single surface layer so that dopants may be infused. Plasma equipment simulations show that high microwave power and high pressure operation are sufficient to generate sufficient dopant radicals with a sufficient non-damaging flux of light ions to the surface to meet functionalization requirements. [11] The properties of the surface layer are very important. Using arsenic doping of silicon as an example, first principles calculations of ternary dopant silicon oxygen system were used to explore the surface film properties. [12] The calculations show that trade-offs between strain relaxation of the amorphous and the evolution of penalty energy as arsenic is introduced to the amorphous silica system creates a preferred stoichiometry film. This is one ingredient to maintaining conformality in the presence of different flux compositions all around a feature. The presentation will close with a look forward at how the new norm in modeling and simulation will involve addition of first-principles (quantum chemistry through device) to the integrated source through feature approach of the last two decades. [1] C. T. Carver et al., ECS J. Sol. St. Sci. Tech., 4 N5005-N5009 (2015) [2] S. M. George, Chem. Rev., 110, 111 (2010) [3] K. J. Kuhn, IEEE Trans. Elec. Dev., 59, 1813 (2012) [4] T. Tatsumi and J. Lee, Jpn. J. Appl. Phys., 53, 06JE06 (2014) [5] K. Ono and M. Tuda, Thin Solid Films, 374, 208 (2000) [6] M. Wang and M. J. Kushner, J. Appl. Phys 107, 023308 (2010) [7] A. Agarwal and M. J. Kushner, J. Vac. Sci. Technol. A 26, 498 (2008) [8] A. Ranjan et al., Proc. SPIE 9428, Advanced Etch Technology for Nanopatterning IV, 94280O (2015) [9] S. King, J. Vac. Sci. Technol. A, 29, 041501(2011) [10] J. Yoshikawa, AVS 60th Annual Symposium, Paper TF+PS-ThM12(2013) [11] H. Ueda et al., J. Appl. Phys., 115, 214904 (2014) [12] P. Ventzek et al., Appl. Phys. Lett., 105, 262102 (2014)
Read full abstract