Silicon Doping by Polymer Grafting: Size Distribution Matters

Abstract

Phosphorus δ-layers in SiO 2 have been prepared by means of poly(methyl methacrylate) (PMMA), terminated with a phosphoruscontaining moiety acting as an anchoring group. In particular, grafting of two P-terminated PMMA samples with M n = 7.5 kg/mol (D̵ = 1.14) and M n = 17.8 kg/mol (D̵ = 1.23) onto 10 nm thick SiO 2 films deposited on Si substrates has been investigated, focusing on the thickness evolution of the brush layer as a function of the processing parameters, that is, annealing temperature and time. Upon removal of the polymer chains and subsequent encapsulation into a SiO 2 matrix, the concentration of phosphorus atoms into the P δ-layers has been monitored by time-of-flight secondary ion mass spectrometry. The effective P dose in the P δ-layer is mainly dictated by the molecular weight of the P-terminated PMMA, and the doping process results are highly reproducible, provided that tight control over the experimental protocol is granted. However, although the grafting density is expected to progressively increase as a function of annealing time with a linear correlation between grafting density and thickness, the measured P dose in the δ-layers is observed to follow the opposite trend. This effect has been accounted for by considering a distortion of the molecular weight distribution of the grafted species with respect to the initial molecular weight distribution of the polymer. The overall picture reveals important information about the mechanism and dynamics governing the "grafting to" process of P-terminated PMMA polymers onto nondeglazed Si substrates.

Short vs. long chains competition during “grafting to” process from melt

“…At the same time, simulations highlight a correlation between the composition of the brush layer and the area occupied by the polymer chains with the reactive chain-end close to the surface, suggesting that the enrichment of short chains during the grafting to process could be not simply related to the number of short and long chains at the interface, but also related to other variables, as the position of the chain term-inals or the accessible reactive areas (see the ESI † for further details). In a recent paper, 50 Perego et al reported clear evidence of preferential grafting of the short polymer chains over the long ones for homopolymers terminated with a reactive phosphorus containing moiety. However, these polymers exhibit a very different evolution of the brush layer composition as a function of time compared to the hydroxy terminated statistical copolymers that were investigated in the present work.…”

“…The combination of the above effects is likely to explain the recently described deviation of the grafting density and the brush thickness from their usual affine and parallel increase as a function of time. 27…”

“…The combination of the above effects is likely to explain the recently described deviation of the grafting density and the brush thickness from their usual affine and parallel increase as a function of time. 27 Understanding the actual "grafting to" mechanism, including the above effects, is important for several cutting-edge applications such as deterministic doping of semiconductors using polymers terminated with dopant containing moieties. In this technology, the amount of dopant atoms to be grafted to the silicon surface, and then driven into the silicon to modify the electrical properties, is controlled by the molecular weight of the dopant bearing polymer employed in the grafting process.…”

Brush Layers of Bioinspired Polypeptoids for Deterministic Doping of Semiconductors

“…16,17 Relatively good control over the dopant atom concentration was demonstrated, 16 and efficient injection and activation of the dopant in the semiconductor substrate were attained. 18,19 However, recent investigations revealed that the "graf ting to" reaction is much more complex 20 than usually believed with a shift from a diffusion-controlled to a mechanochemical driven mechanism. 21 Moreover, partitioning by molecular weight takes place at interfaces in which the lower molecular weight species are preferably incorporated into the polymeric brush.…”

“…22−25 Depending on the specific chemistry of the "grafting to" reaction, this preferential incorporation operates also in the case of polymeric dopants with dispersity Đ < 1.1, featuring relatively narrow molecular weight distributions. 20 These effects limit the capability to effectively control the number of polymer chains in the brush layer, reducing at the end the control on the dose of dopant atoms in the source.…”