Food Dehydration & Pharmaceutical Stability, GAB Moisture Sorption Isotherm
GAB Food Moisture Sorption Isotherm (M0, C, K) Formula
Calculate Effective Multilayer Activity K*aw, GAB Equilibrium Moisture Content (% dry basis), Uncorrected BET Moisture Prediction (%), and High-aw BET Divergence Error (%).
Why classic 2-parameter BET breaks down above aw = 0.45 in dehydrated foods—and how GAB’s third parameter K extends accuracy to aw = 0.90!
Freeze-dried foods, milk powders, pet kibble, and pharmaceutical tablets store shelf-life stability on their **Moisture Sorption Isotherm** relating Dry-Basis Equilibrium Moisture `M` (`% db`, g water / `100` g dry solids) to **Water Activity `a_w`** (`0 < a_w < 1`). Classic 2-parameter **BET theory** (`M0` monolayer moisture, `C` Guggenheim energy constant) assumes multilayer water molecules above the monolayer have the exact same heat of liquefaction as pure bulk water—blowing up hyperbolically toward infinity above `a_w ≈ 0.45`! The **Guggenheim-Anderson-de Boer (`GAB`) Isotherm** (adopted by European COST 90 & AOAC food engineering standards) introduces a third thermodynamic parameter **`K` (`0.70 <= K < 1.00`)** accounting for structured multilayer sorption enthalpy, scaling `a_w` to **`K * a_w`**: **`M_{GAB} = (M_0 * C * K * a_w) / ((1 - K * a_w) * (1 - K * a_w + C * K * a_w))`** (`% db`)!
Verify on Spray-Dried Skim Milk Powder at `25 °C`—Storage_Water_Activity `a_w` = 0.65 (B2), GAB_Monolayer_Moisture `M_0` = 5.80% db (B3), Guggenheim_Energy_Constant `C` = 14.5 (B4), and Multilayer_Factor `K` = 0.84 (B5): Scaled Activity `K * a_w` equals `0.546` (E2), True GAB Equilibrium Moisture equals `12.08%` db (E3), whereas Uncorrected BET (`K = 1.0`) predicts `15.98%` db (E4)—a `+3.90%` db overprediction (E5)!
| GAB vs BET Sorption Isotherm KPI (M0=5.8% db, C=14.5, K=0.84, aw=0.65 vs 0.82) | Formula | Verified Output |
|---|---|---|
| Effective Multilayer Activity K*aw | Multilayer_K * Water_Activity_aw | 0.546 (0.689 @ humid aw=0.82) |
| GAB Equilibrium Moisture (% db) | (M0 * C * K * aw) / ((1-K*aw)*(1-K*aw+C*K*aw)) | 12.08% db (18.07% db @ humid aw=0.82) |
| Uncorrected BET Moisture (% db) | (M0 * C * aw) / ((1 - aw) * (1 - aw + C * aw)) | 15.98% db (31.74% db blow-up @ aw=0.82!) |
| BET Overprediction Error (% db) | Uncorrected_BET_Pct - GAB_Moisture_Pct | +3.90% db (+13.67% db error @ aw=0.82!) |
3-Parameter GAB Multilayer Sorption Isotherm vs 2-Parameter BET Limit Formula
Validate `Water_Activity a_w` in `(0.02, 0.95]`, `Monolayer M_0 > 0.2%`, `Guggenheim C > 0.5`, and `Multilayer K` in `[0.50, 1.00]`!
=LET(aw_f, B2, mz_p, B3, cg_k, B4, km_f, B5, ka_w, km_f * aw_f, mg_p, (mz_p * cg_k * ka_w) / ((1 - ka_w) * (1 - ka_w + cg_k * ka_w)), mb_p, (mz_p * cg_k * aw_f) / ((1 - aw_f) * (1 - aw_f + cg_k * aw_f)), IF(OR(aw_f <= 0.02, aw_f > 0.95, mz_p <= 0.2, cg_k <= 0.5, km_f < 0.5, km_f > 1), "Invalid Isotherm Input", ROUND(mb_p - mg_p, 2)))Tropical Packaging Permeation (aw = 0.82) BET Asymptote Divergence Check (Set Water Activity B2 = 0.82)
Expose the powder pouch to `82%` Equilibrium Relative Humidity (`a_w` B2 = 0.82): GAB Equilibrium Moisture E3 rises realistically to `18.07%` db (`K * a_w = 0.689` in E2), while `1 / (1 - a_w)` drives Uncorrected BET E4 to an unphysical `31.74%` db (`+13.67%` db error in E5)!
=IF(ISTEXT(E5), E5, LET(ka_w, B5 * B2, ROUND((B3 * B4 * ka_w) / ((1 - ka_w) * (1 - ka_w + B4 * ka_w)), 2)))Try the example
Use the original values or change them to test your understanding.