Cryogenic plasma deep-etching for silicon sub-micron structures was studied with the use of modified poly(styrene) (PS) perforated masks obtained from laterally phase separated PS and poly (lactic acid) PLA blend thin films. PS mask was stained by heavy metal (ruthenium) or transferred to an intermediate hard mask (silicon oxide). For the stained mask, optimization of standard STiGer cryogenic plasma etching process led to etched Si cavities with minimal defects at rate of 0.8 μm/min but within a limited depth (~1.4 μm). For intermediate hard mask, optimized STiGer etching process was used in order to improve the reproducibility and to obtain the deeply etched features up to 10 μm depth with minimal defects. A higher etch rate of around 1.2 μm/min was achieved.
In this paper, the etched surface states of GaN using two different etching techniques are reported. In Cl 2 /Ar ICP based plasma, two types of defect were identified: cavities and columns. A strong decrease of the cavity diameter and density on etched surface was observed with the addition of CHF 3 in the chemistry. SF 6 addition instead of CHF 3 leads to an etched surface free of defects with low GaN etch rate similar to that obtained with IBE technique. XPS and AFM measurements revealed the importance of fluorine species which leads to the formation of a Ga x F y passivation layer on the GaN etched surface.
Cryogenic etching processes were successfully applied to ultra-low K (ULK) material for
interconnect applications in back-end-of-the-line part of advanced CMOS technology. The objective
of our experiments is to minimize the plasma induced carbon depletion. The effect of the wafer
cooling is clearly evidenced by the characterization data obtained by FTIR, ellipsometry, mass
spectrometry and SEM after plasma processes at different wafer temperatures.
The increasing development of Micro-ElecroMechanical Systems and the Lab-on-Chip introduces new challenges for deep silicon etching, where the defects such as undercut and crystallographic effect are not tolerated for these systems [1]. Two different kinds of deep etching process are used to realize such structures. First, it is the Bosch process which uses a succession of etching and deposition steps at ambient temperature. In this process, etching is performed with SF6 plasma and deposition with C4F8 plasma. After many steps of etching and deposition, deep and vertical etch profiles are achieved. Although, it is used often, the Bosh process presents many drawbacks such as scalloping effect, low etch rate, contamination, and process shift [1]. The second is the deep cryo-etching which is now a good alternative to the widely used Bosch process in many applications. It is based on SF6/O2 chemistry and the cooling of the substrate to cryogenic temperatures as low as –100°C. In this process the passivation layer SiOxFy is formed on the sidewalls, while silicon is etched at the hole bottom. It provides smooth profiles and high etch rates [2], [3]. However cryo-etching process is sensitive to temperature shifts, which can make etching or passivation dominate. In most cases, that leads to defects such as black silicon, undercut and bowing [4]. Crystallographic effect in cryo-etching process is characterized by Si (111) facets near the footprint of structures on Si(100) wafers as shown in Fig.1. However, this effect is absent on Si(111) wafers [5]. In this paper, the crystallographic effect and holes profiles are investigated by varying the process parameters such as chuck temperature and bias voltage. 2. Results and discussion
Cryogenic deep etching of silicon is investigated using SO2 for passivating the sidewalls of the etched features. The passivating efficiency of SO2 in a SF6/SO2 inductively coupled plasma is assessed comparatively with the traditional SF6/O2 chemistry by means of mass spectrometry and optical emission spectroscopy diagnostics. Emphasis is placed on the evolution of the density of various neutral species (e.g. SiF4, F, O, SOxFy, SFx). These measurements allow us to determine the SO2/SF6 and O2/SF6 gas flow ratios above which a passivation layer forms and inhibits silicon etching. Furthermore, different reaction schemes are proposed to explain the variations in relative densities measured for the two plasma chemistries. In SF6/SO2 plasmas, surface reactions involving SOF and SO2 species with F radicals are favoured, providing a greater number of SOF2 and SO2F2 molecules in the gas phase. In SF6/O2 plasmas, a higher rate of O radicals available for reacting with SFx species can account for the greater concentration in SOF4 molecules. However, these trends are significant for high passivating gas concentrations only. This is consistent with the similar etch results obtained for both chemistries when etching silicon at cryogenic temperatures with a low percentage of passivating gas.
In this Erratum, the authors revisit the “Results and discussion” and “Conclusion” parts of their previous article [Phys. Status Solidi A 206, 2000 (2009)] after correction of one data point for the incorporation efficiency η of phosphorus plotted in Fig. 2 of the article. Figure 2 in the article by Frangieh et al. [Phys. Status Solidi A 206, 2000 (2009)] has to be replaced by the following corrected figure: Plot of the phosphorus incorporation efficiency, η, as a function of the [C*]/[H2] ratio. Within the Section 3, the second and third paragraphs should read as follows: The phosphorus concentration, [P], reached in this study is in the range of 4 × 1017–2 × 1019 P/cm3. The incorporation efficiency of phosphorus (P) is defined by: In Section 4, the second sentence has to be corrected as follows: Phosphorus is successfully incorporated (up to 2 × 1019 P/cm3) with an incorporation efficiency up to ∼0.026%.
