`
`1.2 Pre-history
`
`Two primary technologies which preceded and underlie transparent elec-
`tronics are briefly overviewed. These topics are transparent conductive
`oxides (TCOs) and thin-film transistors (TFTs).
`
`1 Introduction
`
`ent electronics ‘killer apps’ are admittedly either not yet well-defined or
`are presently unrealizable due to currentlimitations in transparent electron-
`ics or in a requisite auxiliary technology. However, this topical ordering
`inversion is meantto be intentionally provocative. Since transparent elec-
`tronics is a nascent technology, we believe that its development will be
`most rapidly and efficiently accomplished if it
`is strongly application-
`driven, andif it is undertaken in a parallel fashion in which materials, de-
`vices, circuits, and system development are pursued concurrently. Hope-
`fully, such a product-driven concurrent developmentstrategy will lead to
`rapid technology assessment, the identification of new and most-likely un-
`expected applications, and an expeditious commercial deployment ofthis
`technology.
`
`1.2.1 Transparent conducting oxides (TCOs)
`
`TCQs constitute an unusual class of materials possessing two physical
`properties - high optical transparency and high electrical conductivity -
`that are generally considered to be mutually exclusive (Hartnagel et al.
`1995). This peculiar combination of physical properties is only achievable
`if a material has a sufficiently large energy band gap so that it is non-
`absorbingor transparent to visible light, i.e., > ~3.1 eV, and also possesses
`a high enough concentration ofelectrical carriers, i.e., an electron or hole
`concentration > ~L0'? cm”, with a sufficiently large mobility, > ~l em? V"
`‘g' that the material can be considered to be a ‘good’ conductorof elec-
`tricity.
`
`The three most common TCOsare indium oxide In2O3, tin oxide SnOz,
`and zine oxide ZnO, the basic electrical properties of which are summa-
`rized in Table 1.1. All three of these materials have band gaps abovethat
`required for transparency acrossthe full visible spectrum.
`
`Note that although the TCQslisted in Table 1.1 are considered to be
`‘good’ conductors from the perspective of a semiconductor, they are actu-
`ally very poor conductors compared to metals. For example, the conduc-
`
`sce
`
`
`
`1,2 Pre-history 3
`
`tivities of tungsten W, aluminum Al, and copper Cu, are approximately
`100,000, 350,000, and 600,000 S cm", indicating that the best In3O; con-
`ductivity (for indium tin oxide or ITO)is about a factor of 10 to 60 lower
`than that of a typical integrated circuit contact metal. The low conduc-
`tance of TCOs compared to metals has important consequences for both
`TCO and transparent electronics applications, some of which are explored
`in this book. The theoretical absolute limit of the conductivity for a TCO
`has been estimated to be 25,000 $ cm” (Bellingham 1992).
`Table 1,1, Electrical properties of common transparent conducting oxides
`(TCOs) Conductivities reported are for best-case polycrystalline films,
`Material
`Bandgap
`Conductivity
`Electron
`Mobility
`(eV)
`(Sem)
`concentration
`(em? V"!
`
`—oo (cm)
`s')
`n,Q,
`3.75
`10,000
`>107!
`35
`ZnO
`3.35
`8,000
`>17!
`20
`SnO,
`3.6
`5,000
`>107
`15
`
`Returning to Table 1.1, notice that all three of the TCOs included in this
`table are n-type, i.e., conductivity is a consequence of electron transport,
`and that the electron carrier concentration is strongly degenerate, i.e., the
`electron density exceeds that of the conduction effective band density of
`states by an appreciable amount (Pierret 1996; Sze and Ng 2007). All of
`the well-known and commercially relevant TCOsare n-type. p-type TCOs
`are a relatively new phenomenon and their conductivity performance is
`quite poor compared to that of n-type TCOs. To a large extent, the poor
`conductivity of p-type TCOs is due to the very low mobility of these mate-
`tials, typically less than ~1 cm? V's", compared to mobilities in the range
`of ~10-40 cm’ V's"! for n-type TCOs.
`
`The n-type mobilities indicated in Table 1.1 are quite small compared to
`those representative single crystal silicon materials and devices, which
`range from ~250-1,500 cm? V's. However, this mobility comparison be-
`tween TCOsandsingle crystal silicon is a bit misleading since single crys-
`tal silicon mobility is not usually specified at doping concentrations as
`large as those typical of TCOs.
`In fact,
`it is reported that single crystal
`silicon mobility is independent of doping concentration above ~10'° cm?,
`with an electron mobility of ~90 cm? V's’! and a hole mobility of ~50 cm?
`v's! (Baliga 1995). A low mobility at high carrier concentrationsis, to a
`large extent, a consequence of intense ionized impurity scattering associ-
`ated with high doping concentrations (Hartnagelet al. 1995).
`
`SEL 2004
`Bluehouse v. SEL
`IPR2018-01405
`
`SEL 2004
`Bluehouse v. SEL
`IPR2018-01405
`
`