Thin film growth models with long surface diffusion lengths

To, Tung B.T.; Sousa, Vitor B. de; Reis, F. D. A. Aarão

doi:10.1016/j.physa.2018.07.024

Cited by 17 publications

(16 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For low SiO 2 thickness, conformal coverage is greatly achieved; the largest structure height of 700 nm shows a smoothening effect: the structure height of the CIGSe layer is less than of the Al back contact. Surface smoothening can be understood considering adatom surface diffusion [ 22 ] and re‐evaporation. [ 23 ] The smoothening has to be considered in optical simulations.…”

Section: Resultsmentioning

confidence: 99%

Ultrathin CIGSe Solar Cells with Integrated Structured Back Reflector

et al. 2020

View full text Add to dashboard Cite

Reducing the thickness of thin-film solar cells reduces the total cost of ownership of solar module production. Theoretically, a 500 nm Cu(In,Ga)Se 2 (CIGSe) solar cell allows more than 20% efficiency. [1-4] However, this requires a very low back contact recombination velocity and advanced optical light management. The standard molybdenum back contact is generally attributed with a high recombination velocity [5-7] and has a poor optical reflectivity. [8] In ref. [9], a ZrN layer on top of molybdenum having a reflectivity of 60% was used with the result of increased quantum efficiency at long wavelengths. An even higher reflectivity and higher gain in photocurrent could be achieved by an Au back reflector applied by post-deposition recontacting after removing the Mo layer. [10] In addition to cost aspects, Au is not suitable for direct deposition of CIGSe thin films at high temperatures due to the chemical reactivity with the chalcogen species during CIGSe deposition. The latter is also true for other highly reflective metals such as Ag, Cu, and Al. [11] Instead, these metals have to be encapsulated by a transparent and conductive diffusion barrier layer. Bissig et al. investigated an Al/InZnO back contact for solar cells with an absorber thickness of 2.1 μm and were able to increase the short-circuit density up to 1.4 mA cm À2 compared with a sample with Mo back contact, despite the relatively high absorber thickness. [12] According to former work on transparent back contacts, the transparent conductor indium-tin-oxide (ITO) is a good candidate as a transparent back contact [13] and as a diffusion barrier. [14] Recently, ref. [15] showed an experimental gain in short-circuit current density of 4.9 mA cm À2 in a structure Mo/Ag/ITO/0.5 μm CIGSe with respect to the reference structure of Mo/0.5 μm CIGSe. Al/ITO-based solar cells can be prepared up to at least 600 C without Al diffusion inside the absorber while providing high reflectivity at long optical wavelengths. [16] This raises the question of the electronic properties of the ITO/CIGSe contact. Keller et al. investigated 650 nm CIGSe bifacial solar cells with a hydrogen-doped In 2 O 3 back contact [17] and quantified the back-contact recombination velocity to the range of 10 7 cm s À1. This value dropped by application of an Al 2 O 3 þ NaF precursor layer stack as confirmed by a strongly increased rear-side external quantum efficiency (EQE) at short wavelengths. Unfortunately, the 5 nm Al 2 O 3 conformal layer reduces the fill factor (FF). [17] Although a flat optical reflector enhances the long-wavelength EQE upon front-side illumination to some extent, light scattering is required in addition to approach the Yablonovitch limit. [18] Yin et al. used ITO as transparent back contact and applied both dielectric SiO 2 nanoparticles on top of ITO and an external Ag mirror behind the substrate glass achieving a very high short-circuit current density of 32.4 mA cm À2 with an absorber thickness of only 390 nm. [19] This was an important step toward a structured an...

show abstract

Section: Resultsmentioning

confidence: 99%

Ultrathin CIGSe Solar Cells with Integrated Structured Back Reflector

et al. 2020

View full text Add to dashboard Cite

show abstract

“…As pointed out above and stressed recently in Refs. [53,58], the existence of this kind of simplified model reproducing the surface features of a more realistic and complex one (the IACV model here) is very important since this allows us to investigate regimes of the latter model, which would not be computationally accessible in a feasible amount of time, through simulations of the former one.…”

Section: Discussionmentioning

confidence: 99%

Height fluctuations in homoepitaxial thin film growth: A numerical study

Carrasco

2020

Phys. Rev. Research

View full text Add to dashboard Cite

We report on the investigation of height distributions (HDs) and spatial covariances of two-dimensional surfaces obtained from extensive numerical simulations of the celebrated Clarke-Vvedensky (CV) model for homoepitaxial thin film growth. In this model, the effect of temperature, deposition flux, and strengths of atom-atom interactions are encoded in two parameters: the diffusion to deposition ratio R = D/F and ε, which is related to the probability of an adatom "breaking" a lateral bond. We demonstrate that the HDs present a strong dependence on both R and ε, and even after the deposition of 10 5 monolayers (MLs) they are still far from the asymptotics in some cases. For instance, the temporal evolution of the HDs' skewness (kurtosis) displays a pronounced minimum (maximum), for small R and ε, and only at long times it passes to increase (decrease) toward its asymptotic value. However, it is hard to determine whether they converge to a single value or different nonuniversal ones. For large R and/or ε, on the other hand, these quantities clearly converge to the values expected for the Villain-Lai-Das Sarma (VLDS) universality class. A similar behavior is observed in the spatial covariances, but with weaker finite-time effects, so that rescaled curves of them collapse quite well with the one for the VLDS class at long times. Simulations of a model with limited mobility of particles, which captures some essential features of the CV model in the limit of irreversible aggregation (ε = 0), reveal a similar scenario. Overall, these results point out that the study of fluctuations in homoepitaxial thin films' surfaces can be a very difficult task and shall be performed very carefully, once typical experimental films have 10 4 MLs, so that their HDs and covariances can be in the realm of transient regimes.

show abstract

“…A noteworthy one is the Clarke-Vvedensky (CV) model [8][9][10][11], in which the adatom diffusion rates follow Arrhenius laws, with energy barriers depending on the local number of bonds. Recently, the scaling properties of a limitedmobility model were compared with the CV model [12], discussing the effects of memory (non-Markovianity) and probabilities of adatom detachment from terrace steps. It was observed that many central features of thermally activated models can be captured with their limited mobility versions [12].…”

Section: Introductionmentioning

confidence: 99%

“…Recently, the scaling properties of a limitedmobility model were compared with the CV model [12], discussing the effects of memory (non-Markovianity) and probabilities of adatom detachment from terrace steps. It was observed that many central features of thermally activated models can be captured with their limited mobility versions [12]. * eluis@ufba.br † thiagoaa@ufba.br ‡ silviojr@ufv.br § randrade@ufba.br…”

Section: Introductionmentioning

confidence: 99%

Local roughness exponent in the nonlinear molecular-beam-epitaxy universality class in one dimension

et al. 2019

View full text Add to dashboard Cite

We report local roughness exponents, α loc , for three interface growth models in one dimension which are believed to belong the non-linear molecular-beam-epitaxy (nMBE) universality class represented by the Villain-Lais-Das Sarma (VLDS) stochastic equation. We applied an optimum detrended fluctuation analysis (ODFA) [Luis et al., Phys. Rev. E 95, 042801 (2017)] and compared the outcomes with standard detrending methods. We observe in all investigated models that ODFA outperforms the standard methods providing exponents in the narrow interval α loc ∈ [0.96, 0.98] consistent with renormalization group predictions for the VLDS equation. In particular, these exponent values are calculated for the Clarke-Vvdensky and Das Sarma-Tamborenea models characterized by very strong corrections to the scaling, for which large deviations of these values had been reported. Our results strongly support the absence of anomalous scaling in the nMBE universality class and the existence of corrections in the form α loc = 1 − of the one-loop renormalization group analysis of the VLDS equation.

show abstract

Thin film growth models with long surface diffusion lengths

Cited by 17 publications

References 39 publications

Ultrathin CIGSe Solar Cells with Integrated Structured Back Reflector

Ultrathin CIGSe Solar Cells with Integrated Structured Back Reflector

Height fluctuations in homoepitaxial thin film growth: A numerical study

Local roughness exponent in the nonlinear molecular-beam-epitaxy universality class in one dimension

Contact Info

Product

Resources

About