Delayed ettringite formation: a CEM I 52.5 N HTS from 5 to 80 °C and back

Before this page

CEM I 52.5 N HTS and CEM II/A-L 42.5 R from 0 to 60 °C, the paste and the temperature at which monosulfate replaces ettringite.

A paste heated above about 50 °C loses ettringite to monosulfate, the sulfate it held going to the solution; cooled, it forms ettringite again, in a hardened paste whose pores were not made for it: the delayed ettringite formation of steam-cured concrete. Lothenbach et al. (2007) hydrated the sulfate-resisting CEM I 52.5 N HTS (SRPC) at w/c 0.4 at 5, 20 and 50 °C and measured its pore solution and its phases. This page computes the paste of the companion paper Lothenbach et al. (2008), at the degree of hydration 150 days give it, at each temperature, then heated to 80 °C and brought back to 20 °C.

using ChemistryLab, DynamicQuantities, Printf, Plots
include(joinpath(pkgdir(ChemistryLab), "scripts", "delayed_ettringite.jl"))

cs = l08t_system()
paste = Dict(20.0 => de_state(20.0; cs))
for T in (5.0, 50.0, 80.0)
    paste[T] = de_state(T; start = paste[20.0].state, cs)
end
println(" T (°C)   SO4 (mmol/L)   OH (mol/L)   ettringite   monosulfate  (g per 100 g of cement)")
for T in (5.0, 20.0, 50.0, 80.0)
    o = de_observables(paste[T])
    @printf("%6.0f %8.3f / %5s %7.3f / %5s %10.2f %13.2f\n", T, 1000o.SO4,
            isnan(de_measured("SO4", T)) ? "-" : @sprintf("%.2f", 1000de_measured("SO4", T)),
            o.OH, isnan(de_measured("OH-", T)) ? "-" : @sprintf("%.2f", de_measured("OH-", T)),
            o.ettringite, o.monosulfate)
end

 T (°C)   SO4 (mmol/L)   OH (mol/L)   ettringite   monosulfate  (g per 100 g of cement)
     5    0.029 /  0.85   0.096 /  0.16      11.70          0.00
    20    0.107 /  2.20   0.095 /     -      11.70          0.00
    50    1.034 / 13.00   0.087 /  0.19      11.70          0.00
    80   15.525 /     -   0.065 /     -       7.96          5.43

Ts = [5.0, 20.0, 50.0, 80.0]
fig = plot(; yscale = :log10, xlabel = "temperature (°C)", ylabel = "dissolved sulfur (mmol/L)", legend = :bottomright,
           size = (700, 400), left_margin = 5Plots.mm, bottom_margin = 5Plots.mm,
           title = "SRPC, w/c 0.4, at the degree of hydration of 150 days")
plot!(fig, Ts, [1000de_observables(paste[T]).SO4 for T in Ts]; lw = 2, marker = :circle, label = "here")
measured_T = [T for T in Ts if !isnan(de_measured("SO4", T))]
scatter!(fig, measured_T, [1000de_measured("SO4", T) for T in measured_T]; color = :black, marker = :diamond,
         label = "measured, Lothenbach et al. (2007)")
fig

Here / measured at 150 days. The sulfate of the solution rises with the temperature, as measured, but from far lower: 13 to 30 times below the measurement, with a rise of 35 from 5 to 50 °C where it was 15. The hydroxide is half the measured one, the sodium a quarter of it and the potassium two fifths. At 50 °C the paste holds no monosulfate yet, the transition of this calculation being at 53 °C (the paste against temperature), where the measured paste held some.

At 80 °C a third of the ettringite has given way to monosulfate. Brought back to 20 °C, the same budget gives the same equilibrium as before the heating:

back = de_state(20.0; start = paste[80.0].state, cs)
o, o20 = de_observables(back), de_observables(paste[20.0])
@printf("ettringite %.2f g, monosulfate %.2f g, as before the heating: %.2f g and %.2f g\n",
        o.ettringite, o.monosulfate, o20.ettringite, o20.monosulfate)
println("ettringite declared from ", temperature_range(L08T_DB["ettringite"])[1], " to ",
        temperature_range(L08T_DB["ettringite"])[2], " K")

ettringite 11.70 g, monosulfate 0.00 g, as before the heating: 11.70 g and 0.00 g
ettringite declared from 273.15 to 333.15 K

stages = ["5 °C", "20 °C", "50 °C", "80 °C", "back to 20 °C"]
obs = [de_observables(paste[T]) for T in Ts]
push!(obs, o)
xs = collect(eachindex(stages))
fig = bar(xs .- 0.2, [x.ettringite for x in obs]; bar_width = 0.4, label = "ettringite", color = :steelblue,
          xticks = (xs, stages), ylabel = "g per 100 g of cement", legend = :topleft, size = (750, 400),
          ylims = (0, 1.25 * maximum(max(x.ettringite, x.monosulfate) for x in obs)),
          left_margin = 5Plots.mm, bottom_margin = 5Plots.mm, title = "The paste heated to 80 °C and cooled")
bar!(fig, xs .+ 0.2, [x.monosulfate for x in obs]; bar_width = 0.4, label = "monosulfate", color = :darkorange)
fig

The equilibrium is reversible: the ettringite comes back, which is what makes the delayed formation possible, and the calculation cannot say where or when. At 80 °C the heat capacity of ettringite is extrapolated, Cemdata18 (Lothenbach et al., 2019) declaring it to 60 °C only (temperature_range).