Electrical Transport Mechanisms in Ensembles of Silicon Nanocystallites

Abstract

While the optical [1][2][3] properties of various ensembles of individual Si nanocrystallites (NCs) and the memory-charge storage (CS) characteristics of two-dimensional (2D) arrays of Si NCs [4] have been investigated by many researchers, relatively little attention was paid to the transport properties of 3D ensembles of such quantum dots (QDs) [5]. The interest in the last systems, however, is expected to follow the new basic physics that it reveals and the potential applications of such systems. Following the reasonable (though still controversial [6]) understanding of granular metals [7,8], where the electrical conduction takes place via metallic particles, additional significant insights into the transport mechanism [5,9] and new phenomena are expected in Si NCs due to the presence of confined levels [10]. In particular, the combination of Coulomb blockade (CB) effects and the quantum confinement (QC) restrictions can yield various resonant tunneling effects [11] as well as various quantum phase transformations [12] that are beyond the scope of this chapter. From the application point of view, one realizes that significant electroluminescence (EL) can come about only as a result of efficient transport in 3D ensembles of Si NCs [4,13,14] that are dense enough to yield strong light emission. Finding the conditions for the optimization of the luminescence and the transport is then the route to achieve efficient Si-based photoelectronic devices [15].In this chapter, we present a review on the electrical transport in the ensembles of Si NCs. Since we are concerned with the transport in 3D ensembles of quantum dots, we will only briefly review the main results obtained from lower dimensional ensembles (i.e., 1D-like [16, 17] or perpendicular to 2D arrays [18][19][20][21]) of Si QDs in order to provide a reference for the effects of the small size of the single Si NCs on the transport in ensembles of larger dimensions. We will see that in spite of many studies of these 0D-like and 2D systems, which are essentially associated with single isolated NCs, and their potential use as nonvolatile memories, only the basics of the Silicon Nanocrystals: Fundamentals, Synthesis and Applications. Edited by Lorenzo Pavesi and Rasit Turan

Glassy-like Transients in Semiconductor Nanomaterials

“…The glassy feature of memory is clearly exhibited in the figure by the fact that the I(t) of each subsequent pulse picks up where the I(t) of the previous pulse left off. The times involved here (hours) are much longer than any microscopic (recombination-like [ 20 ]) or macroscopic (circuit-RC-like [ 40 ]) time constants of the sample (less than seconds), thus confirming the sluggishness of the relevant transport process during and past the pulse application. Having these memory and sluggishness features of the glassy behavior, we have checked the fulfillment of the most characteristic feature of the glassy behavior, the aging.…”

“…In particular, for the electrical systems, the “sluggishness” in manifested by a characteristic time constant τ of the rise and/or the decay of σ(t) which is much longer than any experimental macroscopic or intrinsic microscopic time of the system such as carrier scattering or recombination times [ 20 ] (which are less than a msec. in Si NCs [ 39 , 40 ]).…”

“…In glasses, this can be the distribution of local viscosities [ 25 ], while in nanocomposites, this can be the distribution of the tunneling barrier widths between nearest neighbors or other random potential fluctuations through the system [ 47 , 48 ]. In particular, the latter distributions can be associated with grains’ charging [ 49 , 50 ] and/or level quantization of neighboring nanocrystallites [ 1 , 40 ]. On the other hand, in some nanosystems, the data fit the logarithmic relation [ 33 , 51 ] ln[I(t)/I r ] = A − Bln(t), where A and B are fitting parameters [ 43 ] that enclose more detailed physical information of the system Here, we remark that in glasses the kinetics can vary from stretched exponential [ 24 ] through ln(t)-like to 1/t-like dependencies [ 52 , 53 , 54 , 55 , 56 ].…”

“…The third and more fundamental question is whether an observed SGB in these systems is just an artificial empirical–experiential presentation of experimental results or is it due to some basic physical resemblance between the collective Coulomb mechanism [ 70 , 71 ] and the collective mechanical (say, the viscosity) mechanism that results in SGB in the classical glass [ 24 , 35 , 67 ]. In the work reported here, we tried to respond to these three questions with an experimental study on two types of nano-semiconductors for which the QC-CB-CS effects have been previously observed by others [ 1 , 5 , 6 , 63 ] and by us [ 3 , 40 , 42 ].…”

“…The samples we studied are nano-CdSe photoconductors [ 1 , 76 , 77 ] and the nano-Si-SiO 2 composites [ 3 , 22 , 75 ] on which we conducted optical [ 3 , 9 ], electrical [ 15 , 76 ], and nanospectroscopy [ 11 , 77 ] measurements. In [ 77 ], we detailed our results on nano-CdSe, and in [ 3 , 9 , 40 , 76 , 79 ], we detailed our results on nano-Si-SiO 2 . Here, we mention that such systems have already been shown to result in quite a few optoelectronic applications [ 5 , 13 ], the conspicuous of which are quantum dots [ 50 , 63 , 80 , 81 , 82 , 83 ] and quantum well [ 47 , 48 , 58 , 84 , 85 ] devices.…”