Lecture 2: Cement Hydration Nature of hydrates Thermodynamic prediction of hydrate assemblages Evolution of microstructure Cement Chemistry for Engineers, Cape Town 31 st January 2013
Reaction between water and cement: Increasing solid volume, replaces water Transformation from fluid paste to solid Cement grain water hydrates
alite aluminate ferrite belite NB cement grains are polymineralic
Hydrate Phases
Basic Reactions Alite, C 3 S + H Belite, C 2 S + H Aluminate,C 3 A + 3C$+ H AFt + 2C 3 A + H Ferrite, C 4 AF C-S-H + CH C-S-H + CH AFt (ettringite) AFm Silicates Aluminates Limestone CaCO 3
Phases present 14 month paste from Taylor C-S-H 48% C3S 70% C3A 8% ferrite 7% C2S 15% ett (AFt) 4% CH 14% AFm 11% other 4% pores 16% anhyd 3%
Hydration of calcium silicates : C 3 S and C 2 S calcium hydroxide Hydrated lime portlandite Ca(OH) 2 CH crystalline Hexagonal morphology ~ 15-25% of hydrated paste hydrated calcium silicate C-S-H Nano crystalline multiple morphologies ~ 50-65% of hydrated paste
C-S-H atomic structure and composition meso structure microstructure / morphology
C/S C-S-H: range of compositions 2,50 2,00 1,50 1,00 Flint and Wells at 30 C Taylor at 17-20 C C/S = 0.8 to 1.5 for synthetic preparations 1.7-2 PC pastes 0,50 0,00 0,00 10,00 20,00 30,00 40,00 Source Nonat, Taylor conf CaO mmol/l Lecoq 20 C Thordvaldson 25 C (+expé sursat) Definite phase with reproducible behaviour
C-S-H, analogy with natural mineral Tobermorite Layers of Ca-O Tetrahera of Si-O 4 linked in chains Molecules of water 14Å 1.4nm Ca/Si=0.833
The increase of Ca/Si ratio is attributed to three mechanisms : - the abscence of some bridging tetrahedra (Q1/Q2 increases) H H H H H H H Ca Ca Ca Ca Ca Ca Ca H Ca/Si = 0.72 Ca H H H H H H H H Ca/Si = 0.8 Ca Ca Ca Ca Ca Ca Ca Ca H H H H H H H H Ca/Si = 0.9 Ca Ca Ca Ca Ca Ca Ca Ca Source Nonat, Taylor conf
The increase of Ca/Si ratio is attributed to three mechanisms : Source Nonat, Taylor conf - the abscence of some bridging tetrahedra (Q1/Q2 increases) - the substitution of a part of the protons by calcium ions H H H H H H H H Ca/Si = 0.9 Ca Ca Ca Ca Ca Ca Ca Ca H Ca H H H H H Ca/Si = 1 Ca Ca Ca Ca Ca Ca Ca Ca H Ca Ca Ca H Ca/Si = 1.22 Ca Ca Ca Ca Ca Ca Ca Ca
The increase of Ca/Si ratio is attributed to three mechanisms : - the abscence of some bridging tetrahedra (Q1/Q2 increases) - the substitution of a part of the protons by calcium ions - the presence of Ca-OH regions In presence of portlandite (CH) high Ca/Si ratio: Almost exclusively dimers Almost all terminating Ca, not H
Silicate polymerisation during hydration CL = 2 : dimer CL = 5 : pentamer CL = 8 : octamer Change occurs very slowly, years; faster at higher temperatures Also limited by portlandite presence
S/Ca C-S-H: Si substitution by Al - Al H + 0.5 0.4 0.3 0.2 28 days 1 year 3 years Ettringite Monosulfate - - - - - - 0.1 0.0 0.0 CH 0.1 0.2 0.3 0.4 0.5 0.6 CSH Al/Ca OPC pastes: Al/Ca ~ 0.05 ; Al/Si ~ 0.1 Higher with Al rich SCMs
C-S-H atomic structure and composition meso structure microstructure / morphology
Evidence No long range order intrinsic porosity of 26-28% (Powers) gel porosity - from drying therefore upper limit Scattering experiments (neutron, X-ray) and proton NMR indicate characteristic size of about 4-5 nm
Jennings model
Nano crystalline? ~5nm
Meso structure Two interpretations of nanocrystalline nature: 1. Granules independent blobs 2. Sheets with disorganised structure
Open question General agreement that C-S-H consists of nanocrystalline regions: The main open question is whether they are discrete or linked by sheets Important issues for water transport through C-S-H
C-S-H atomic structure and composition meso structure microstructure / morphology
SEM fracture surface 3 hrs HVEM, wet cell, C 3 S, 24hrs 10 hrs
Two microstructurally distinct forms: Outer or early inner or late
TEM from Richardso
C-S-H summary Atomic level structure fairly well understood: CaO sheets with chains (dimers) of SiO 4 tetrahedra attached Al substitutes for Si, in bridging sites Meso level structure less clear Nanocrystallites or nanocrystalline regions with characteristic scale of about 5nm Microstructure Outer, formed early through solution Inner formed later
Aluminate hydrates ettringite C3A.3C$.H 32 AFt aluminate ferrite tri Possible exchange SO 2 2 4 CO3 Ca 3 Si(OH) 6 (CO 3 )(SO 4 ) 12H 2 O (thaumasite) AFm aluminate ferrite mono [Ca 2 Al(OH) 6 ] + layers Many inter layers ions: 2 SO 4 monsulfate OH C4AH13 AlSi(OH) 8 2 CO 3 Cl stratlingite monocarbonate Friedel s salt hydrogarnet C3AH6 katoite Possible exchange H2 Si
Hydration of aluminate phases : C 3 A C 3 A + water Fast reaction Large plates of hydrates stiffening flash set Calciumaluminate hydrates : C 2 AH 8, C 4 AH 13 : AFm phases then C 3 AH 6
Hydration of aluminate phases : C 3 A C 3 A + CaSO 4 H x _ Setting regulation by sulfates C A 3CSH 26H C A.3CS.H (ettringite) 3 2 3 32 _ 2 2 4 4 3 32 6Ca 2Al(OH) 3SO 4OH 26H Ca A.3CS.H _ Exhaustion of sulfate _ 2C A C A.3CS.H 4H 3C A.CS.H 3 3 32 3 12 AFm phase stablised by solid solution It forms in the presence of anhydrous C 3 A In presence of CaCO 3 (fine limestone) monocarbonate forms instead of monosulfate _ Calcium aluminate monosulfate
C 3 A + water + calcium sulfate
Note Reaction of C 3 A is not blocked by ettringite Reaction controlled by absorption of sulfate ions at reacting sites
Hydration of aluminate phases : C 3 A Cement paste after 2-7 days - local formation of monosulfate inside layer of C-S-H where there is C 3 A available. Ettringite remains on exterior of grain.
2. Hydration Kinetics 3. Microstructure formation
Hydrates form through solution due to difference in solubility of anhydrous compounds and hydrates [conc Y] equilibrium curve of the cement phase condition of equilibrium of the solids (thermodynamics) equilibrium curve of the hydrate [conc X] Path of solution composition resulting from the dissolution of solids precipitation [nucleation and growth] with dissolution
Heat Flow (mw/g) stage 1 stage 2 stage 3 stage 4 Alite reaction 4 3 2 1 0 0 3 6 9 12 15 18 21 24 Age of Specimen (Hours)
Heat Evolution Rate [mw/g] Portland cement: silicate plus aluminate 5 4 Secondary formation of ettringite 3 2 Alite reaction Formation of AFm phase 1 0 0 10 20 30 40 50 Time [h]
Heat flow Initial dissolution Nucleation and growth Induction ~10 m ~3 h ~10 h ~24 h time 1μm 1μm
Heat flow Initial dissolution Nucleation and growth Induction ~10 m ~3 h ~10 h ~24 h time
Formation of Hadley grains (hollow shells) 12hr cement paste: ion thinned section TEM 28 days, SEM For some (as yet unknown reason) the alite dissolves beneath the outer C-S-H, leaving a gap or low density area This hydrates of seems to grow the outside this shell, small hollow shells persist in mature concrete The presence of aluminate seems to be important as these shells are not so clearly separated in the case of C 3 S
Summary Microstructural Development SCRIVENER, 1984
50 mm
Hydration of different phases
belite (%wt) ferrite (%wt) alite (%wt) aluminate (%wt) Ph D Vanessa Kocaba (NANOCEM, CP4) 100 80 100 80 60 40 20 A B C D 60 40 20 0 100 1 10 100 0 100 1 10 100 80 80 60 60 40 40 20 20 0 1 10 100 Time (d) 0 1 10 100 Time (d)
Reaction of Belite
Old concrete, only belite and ferrite apparent
Reaction of ferrite phase XRD shows reaction but in BSE appears unreacted Alumina and calcium leach out leaving Fe relic as hydrous Fe hydroxide of iron rich hydrogarnet
Aluminate phases in hardened cement paste 1 mm Intermixing of ettringite and/or Monosulfate at submicron scale
S/Ca Al/Ca Microprobe analyses 0.6 0.5 0.4 Monosulfo 28 days 1 year 3 years 28 days. Formation of C-S-H, CH ettringite and then monosulfate 2C A C A. 3CS. H 4H 3C ACS.. H 3 3 32 3 12 0.3 Ettringite 0.2 0.1 CH C 0.0 1.7 SH 0.0 0.1 0.2 0.3 0.4 0.5 0.6 (Si+Al)/Ca 0.5 0.4 28 days 1 year 3 years Ettringite 0.3 Monosulfate 0.2 0.1 1, 3 years. Further formation of C-S-H, CH Reaction of ettringite into monosulfate 0.0 0.0 CH 0.1 0.2 0.3 0.4 0.5 0.6 CSH Al/Ca Source thesis Severine Lamberet
Ettringite is commonly found in old concrete through recystallisation Quantity may increase due to carbonation sulfate released as monosulfate converts to moncarbonate This reaction is not damaging
Impact of temperature on microstructural development PhD Xinyu Zhang http://library.epfl.ch/theses/?nr=3725
Heat evolution rate (mw/g) Effect of temperature 30 25 20 55 C 15 10 40 C 20 C 40 C 55 C 5 0-5 0 5 10 15 20 25 30 Time (hours) 20 C
Cumulative heat (J/g) 300 250 200 20 C 40 C 55 C 55 C 40 C Arrhenius equation k = A exp(-e a /(RT)) ln(k) = ln(a) - E a /(RT) 150 100 20 C k A Ea R T rate of reaction constant apparent activation energy gas constant, 8.31 J/K mole temperature in Kelvin 50 0 1 2 3 4 5 6 7 8 910 20 30 Time (hours)
Compressive strength (MPa) Water sorptivity (m sec -1/2 10-6 ) Temperatures does not affect mechanical properties and durability in the same way 55 50 45 40 35 8 7 6 5 5 C 20 C 40 C 60 C 30 25 20 15 10 5 5 C 20 C 40 C 60 C 4 3 2 1 0 0 2 4 6 8 10 12 14 16 18 20 Time (d 1/2 ) strength 0 0 2 4 6 8 10 12 14 16 18 20 Time (d 1/2 ) Water absorption
Why? Lower final degree of hydration at higher temperatures? Different hydration products? Need strength as a function of hydration degree not time
compressive strength (MPa) 60 (w/c=0.5) 50 40 30 20 20 C 40 C 60 C 5 C 10 0 20 40 60 80 100 hydration degree (%)
CSH relative density 20 C 0.45 Due to lower microporosity in C-S-H more capillary porosity at higher temperatures. 90d inner CSH is very thin 0.40 0.35 0.30 60 C 0.25 0.20 0.15 20 C 40 C 60 C 0.10 0.05 0.00 w/c=0.6 w/c=0.35 CEM42. 5 mix1 mix5 mix4 mix2 mix3 CEM52.5 R
C-S-H formed at 90 C Brighter = more dense; less microporosity C-S-H formed at 20 C Darker = less dense; more microporosity Building Materials Analysis 2009 58
Cement Chemistry for Engineers, Cape Town 31 st January 2013 End Lecture 2