Version v1.2 · 2026-10-14
R134a–DMF. Zehioua et al. (2009) correlated 58 measured points (303.30–353.24 K) with a Peng–Robinson + MHV1 + NRTL framework (τ₁₂ = 2250 J/mol, τ₂₁ = −2650 J/mol, α = 0.3), reporting AARD 1.49% and maximum 3.01%. Those parameters do not transfer to a γ–φ formulation (direct transfer underestimates pressures by about 20%); the library carries its own γ–φ NRTL re-fit to the same data, and a cross-laboratory comparison (Han 2011; Cui 2007; Deng 2014) places the inter-laboratory spread at ≈8% at 300 K and ≈20% at 360 K. A second published set (He et al. 2009) is excluded: its printed mixture-enthalpy and VLE equations return non-physical values on recomputation (TE-024).
R134a–DMA. Nezu et al. (2002) published bubble-pressure and enthalpy correlations; the library holds a Python reference implementation. The DMA and DMF columns of the paper's Table 2 carry suspect signs in their B coefficients (the computed deviation sign contradicts the paper's own Fig. 4), while the DEGDME, TrEGDME and TEGDME columns check out (TE-028) [to be verified].
R124–NMP and R124–DMF. Xu et al. (2017) correlated both pairs with a five-parameter NRTL model over 303.15–363.15 K; recomputation reproduces the paper's 53 (NMP) and 60 (DMF) points with AARD 1.25% and 1.17%, matching the reported values. Both pairs are VLE-only (Section 6).
R124–DMAC. Borde, Jelinek & Daltrophe (1997) published a complete correlation set (pure-component vapour pressures, bubble pressure, excess enthalpy, solution enthalpy, density, viscosity). The transcription in the library was verified against the paper's own endpoints and figures; two misprints in the paper were arbitrated and documented, and a thesis transcription that altered two of its coefficient tables is registered (TE-007–TE-009).
R22–DMF and R22–DEGDME. For R22–DMF the library fits bubble pressure to Agarwal's 1982 VLE data (AARD 2.38% on 132 points) and takes solution enthalpy from Fatouh (1993), with −25 °C as the lower boundary. For R22–DEGDME, Ando & Takeshita (1984) published bubble-pressure, heat-capacity and mixing-heat correlations whose paper-reported deviations (0.79%, 0.83%, 0.72%) are reproduced by the library implementation.
Further NRTL sets. Feng et al. (2016) published five-parameter NRTL correlations, τ = (a + b·ln T)/(RT), for HFC161–DMETrEG, HFC152a–DMETrEG and HFC152a–NMP; two printed values are flagged — α = −37.1878 for HFC152a–NMP, and b₂₁ = −99.24 exactly equal to a₁₂ for HFC161–DMETrEG [to be verified].
A binary-refrigerant case, (R32 + R227ea)/DMF (Gao Wanli et al. 2010), combines Wagner equations, a Peng–Robinson equation and UNIFAC; the paper states that no measured VLE for the ternary system has been published and gives no solution-enthalpy model, so the set is registered as having no experimental anchor.
Hydrocarbon pairs are thinner. R600a + POE ISO 7 oil (Neto & Barbosa 2008) was correlated four ways; the authors' Raoult-form reduction has AAD ≈ 5%. For R290–oil, Wang, Jia & Wang (2021) give a p–h–X chart but the full text has not been obtained; a 2018 thesis (Li Keqiao) gives an 11-coefficient rational-polynomial bubble-point fit for R290–NM100P oil (20–120 °C, R² = 0.9993) whose printed form and units await checking [to be verified].
References are numbered in Section 7 of the overview page.