Efficiency limits for photoelectrochemical water-splitting

Atlas use: this paper is cited as evidence by 4 claims and 0 named pathways. Not yet checked against Crossref.

Read in full (open Nature page and PDF, loop-3 pass 46). Ideal case (Table 1: absorption fraction 1, external radiative efficiency 1, catalytic exchange current densities infinite, series resistance 0, shunt resistance infinite; nine assumptions listed — AM1.5G, complete above-gap absorption, detailed-balance dark current with only radiative recombination, ideality 1, negligible catalytic overpotentials, charge-transfer coefficients 0.5, no series resistance, E_rxn = 1.23 V, unity Faradaic efficiency): 'The maximum single and dual junction efficiencies are 30.6% at a bandgap of 1.59 eV, and 40.0% with bandgaps of 0.52 and 1.40 eV, respectively'; the ideal triple junction maximum is 28.3 % — additional junctions add nothing because photovoltage beyond that needed to split water does not raise efficiency. Realistic cases: Real1 (high-performance materials: f_abs 0.9, ERE 0.03, j0,cat {1, 10⁻³} mA cm⁻² cathodic/anodic, r_s 0, r_sh ∞) η_max 15.1 % at 2.05 eV single junction and 28.3 % at 1.59 / 0.92 eV dual junction; Real2 (Earth-abundant: f_abs 0.9, ERE 10⁻⁶, j0,cat {1, 10⁻⁵}, r_s 0.1, r_sh 10, normalised resistivities) 5.4 % at 2.53 eV and 16.2 % at 1.93 / 1.38 eV — the paper's words are 'realistic limiting efficiencies', maxima of parameterised model cases.

Exploreatlas context · claims citing it · cite