Recent studies of cements and concretes by synchrotron radiation crystallographic and cognate methods

CE: KRR QA: Coll: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 Crystallography Reviews, 2015 http://dx.doi.org/10.1080/0889311X.2015.1070260 REVIEW Recent studies of cements and concretes by synchrotron radiation crystallographic and cognate methods Miguel A.G. Aranda∗ ALBA Synchrotron, Ctra, BP1413 km. 3,3, Cerdanyola del Vallès, Barcelona 08290, Spain (Received 22 May 2015; accepted 4 July 2015) The portfolio of available synchrotron radiation techniques is increasing notably for cements and pastes. Furthermore, sometimes the terminology is confusing and an overall picture highlighting similarities and differences of related techniques was lacking. Therefore, the main objective of this work is to review recent advances in synchrotron techniques providing a comprehensive overview. This work is not intended to gather all publications in cement chemistry but to give a unified picture through selected examples. Crystallographic techniques are used for structure determination, quantitative phase analyses and microstructure characterization. These studies are not only carried out in standard conditions but synchrotron techniques are especially suited to non-ambient conditions: high temperatures and pressures, hydration, etc., and combinations. Related crystallographic techniques, like Pair Distribution Function, are being used for the analysis of ill-crystalline phase(s). Furthermore, crystallographic tools are also employed in imaging techniques including scanning diffraction microscopy and tomography and coherent diffraction imaging. Other synchrotron techniques are also reviewed including X-rays absorption spectroscopy for local structure and speciation characterizations; small angle X-ray scattering for microstructure analysis and several imaging techniques for microstructure quantification: full-field soft and hard X-ray nano-tomographies; scanning infrared spectro-microscopy; scanning transmission and fluorescence X-ray tomographies. Finally, a personal outlook is provided. Keywords: synchrotron techniques; quantitative phase analysis; imaging and mapping; binders; cement hydration; cement paste microstructure; C–S–H and C–A–S–H gels Q16 Contents PAGE 1. A very brief introduction to Portland cement and concretes 3 2. A very brief introduction to synchrotron radiation and properties 5 3. Uses of synchrotron radiation for cements, pastes and concretes 7 3.1. Diffraction for characterizing crystalline phases (long-range order) 10 3.2. Techniques for characterizing amorphous phases 10 3.2.1. Total scattering techniques for studying short-range order 10 3.2.2. X-ray absorption techniques for studying short-range order and chemical states 11 3.3. Diffraction for quantifying crystalline phases and in many cases phase evolution(s) 12 3.3.1. Quantitative phase analysis of anhydrous building materials 12 *Emails: [email protected];[email protected] c 2015 Taylor & Francis Techset Composition India (P) Ltd., Bangalore and Chennai, India GCRY1070260.TeX Page#: 46 Printed: 17/7/2015 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 2M.A.G. Aranda 3.3.2. Quantitative phase analysis of hydrating binders 13 3.4. Variable-temperature diffraction to characterize phase evolution(s) and reactions 15 3.4.1. In-situ thermal formation and decomposition studies at moderate temperatures 15 3.4.2. In-situ clinkering studies at very high temperatures 16 3.5. Diffraction under pressure to characterize phases and chemical reactions 16 3.5.1. Chemical reactions at relatively low pressures in capillary cells (hydrothermal conditions) 17 3.5.2. Hydrating reactions at medium pressures and temperatures 18 3.5.3. Phase characterization at very high pressures in diamond-anvil-cells 18 3.6. Small angle X-ray scattering for microstructural characterization of building materials 20 3.7. Imaging building materials at different length scales and with different photon energies 21 3.7.1. Full-field soft X-ray nano-tomography 23 3.7.2. Full-field hard X-ray micro-tomography 24 3.7.3. Scanning synchrotron radiation microscopies 28 3.7.4. Hard X-ray coherent diffraction imaging 34 4. Outlook 39 Acknowledgments 40 Disclosure statement 40 References 40 Nomenclature, Acronyms Cement nomenclature C=CaO, C=CO2,S =SiO2,S =SO2,A =Al2O3,F =Fe2O3,M =MgO, K=K2O, N=Na2O and H=H2O. Therefore, for instance,Ca3SiO5is C3S, CaCO3is CC, and CaSO4.2H2OisCSH2. AFm ‘Al2O3–Fe2O3–mono sulphate’ (set of phases) Aft ‘Al2O3–Fe2O3–tri sulphate’ (aka, ettringite) ASR alkali silica reaction CSA calcium sulphoaluminate cement C–S–H calcium–silicate–hydrate amorphous gel DTA/TGA differential thermal analysis/thermo gravimetric analysis MIP mercury intrusion porosimetry OPC ordinary Portland cement PCE polycarboxylate ether SCM supplementary cementitious materials w/c water-to-cement mass ratio AChX-nCT absorption-contrast hard X-ray nano-computed tomography AChX-µCT absorption-contrast hard X-ray micro-computed tomography ACsX-nCT absorption-contrast soft X-ray nano-computed tomography ACsX-nM absorption-contrast soft X-ray nano-microscopy BCDI Bragg coherent diffraction imaging BL beam line CAT computed axial tomography (aka, computed tomography) 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 Crystallography Reviews 3 CDI coherent diffraction imaging (aka, lensless imaging) CRL compound refractive lenses DAC diamond anvil cell EXAFS extended X-ray absorption fine structure FCDI forward coherent diffraction imaging FoV field of view FZP Fresnel zone plate IR infrared KB Kirkpatrick–Baez (elliptically-bent double-focusing mirrors) PChX-µCT phase-contrast hard X-ray micro-computed tomography PDF pair distribution function PFCDI-nCT ptychographic forward coherent diffraction imaging nano-computed tomography RQPA Rietveld quantitative phase analysis SANS small-angle neutron scattering SAS small-angle scattering SAXS small-angle X-ray scattering SCXRD single crystal X-ray diffraction SEM scanning electron microscopy SFXM scanning fluorescence X-ray microscopy SR synchrotron radiation STXM scanning transmission X-ray microscopy SDXM scanning diffraction X-ray microscopy (aka, synchrotron microdiffraction) SDX-µCT scanning diffraction X-ray micro-computed tomography SXRPD synchrotron X-ray powder diffraction TEM transmission electron microscopy XANES X-ray absorption near edge structure (aka, NEXAFS near-edge X-ray absorption fine structure) XAS X-ray absorption spectroscopy XFEL X-ray free electron laser XRPD X-ray powder diffraction 1. A very brief introduction to Portland cement and concretes The term cement (building material) may be used for almost any type of binder, with chemistry ranging from fully organic to totally inorganic. The oldest organic ‘cements’ were made from plant extracts like starch. The oldest pure inorganic cements were based on muds and subsequently on gypsum plasters. However, since the development of Portland cement about 180 years ago (http://en.wikipedia.org/wiki/Portland_cement), it has become the dominant binder used in concrete (http://en.wikipedia.org/wiki/Concrete) for construction. Concrete is a hierarchically complex hardened composite material formed from the mixing of water and aggregates (both fine and coarse) with Portland cement. Some other additions and admixtures may be added depending upon the type of concrete.[1] Portland cements are used as poorly soluble but highly reactive powders in water. The initial suspension, named as fresh cement paste, firstly undergoes setting and on hardening yields the cementitious matrix. The use of water for achieving the hardening results in the term ‘hydraulic binders’, to highlight that the reaction of cement with water yields the concrete which can harden even underwater. 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 4M.A.G. Aranda Colour online, B/W in print Figure 1. Embodied energy for different typesof materials normalized to the unit volume. Reprinted from Materials and the Environment: Eco-informed Material Choice, Second Edition, M. F. Ashby, Chapter 6: Eco-data: Values, sources, precision, Figure 6.11. Copyright (2013), with permission from Elsevier. Praised for its versatility, durability, fire-resistance, on-site casting and economic value, Portland concrete is receiving recognition also for its relatively low embodied energy compared to other building materials (see Figure 1). Furthermore, because of its use of local materials, by which there is reduction in energy and pollution costs associated with material transport, it is Q1 also an important asset. Annual worldwide Portland cement production is approaching 3 Gt.[2] World yearly consumption of concrete is well over 6 km3;its magnitude may be, perhaps, better perceived by referring this value to every person in the planet: 2.5 tonnes per person per year. This is just second only to our consumption of fresh water. However and in spite of its universal use, Portland cements are one of the most environmentally contentious materials. Worldwide production of cement accounts for approximately 6% of the total anthropogenic CO2production.[3] Research in cements and concretes has many facets ranging from the understanding of phase assemblage and microstructure to ensure extended durability, to reduce the CO2footprint of cements to mitigate the current man-increased green-house effect. An overview of the research carried out in cements can be gained from the scientific programme of the International Congress on the Chemistry of Cements which takes place every four years with the last edition taking place in October-2015 at Beijing (www.iccc2015beijing.org/). Portland cement is manufactured by grinding the Portland clinker with the setting regulator that may be close to 4 wt% of gypsum. A typical Portland clinker chemical composition is close to 67wt% CaO, 22 wt% SiO2,5wt%Al 2O3,3wt%Fe 2O3and 3wt% of minor components. With this elemental chemistry, OPCs usually contains four major phases: 50–70 wt% alite, Ca3SiO5 or C3S, 15–30 wt% belite, Ca2SiO4or C2S, 5–10 wt% tricalcium aluminate, Ca3Al2O6or C3A and 5–15 wt% tetracalcium aluminoferrite, Ca4Al2Fe2O10 or C4AF. Most cement compounds are not pure stoichiometric phases but they (may) incorporate many ions as extensively discussed in classical papers and books.[4] 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 Crystallography Reviews 5 Finally, some crystalline and amorphous hydrated phases should also be mentioned. AFm are a set of phases with crystalline layered structures. The AFm phase refers to a family of hydrated calcium aluminates based on the hydrocalumite structure, Ca4Al2(OH)12.[Cl(OH)].6H2O. The archetype AFm phase is C3A.CaSO4.12H2OorCa 4Al2(OH)12.[SO4].6H2O, known as Kuzelite, but Al can be partly replaced by Fe and SO2− 4can be partly or fully replaced by OH−,Cl −, CO2− 3and several other anions.[5,6]Inasimilarway,AFt are a set of crystalline tridimensional framework compounds. By far the most common AFt phase is ettringite which has the stoichiometry C3A.3CaSO4.32H2O that can also be written as Ca6Al2(SO4)3(OH)12.26H2O. On the other hand, the principal binding reaction product of cement hydration is amorphous calcium– silicate–hydrate (C–S–H) gel, one of the most complex of all gels.[7] Saturated C–S–H gel has the approximate formula (CaO)1.7SiO2(H2O)4, including liquid water between the particles, but the Ca/Si ratio and the water content evolves with time and it depends upon the composition of the starting binder. 2. A very brief introduction to synchrotron radiation and properties Current synchrotron light sources are a particular type of circular particle accelerator designed and built to produce electromagnetic radiation of outstanding properties. These laboratories are user-oriented (both academic and industrial) Large Facilities where staff of very different backgrounds (photon scientists: physicists, chemists, geologists, biologists, physicists, etc.; and supQ2 port scientists and technologists: accelerator physicists, mechanical/vacuum/computing/control engineers, technicians, etc.) work together with the final goal of offering the best possible service to the users. A synchrotron light source usually contains three types of electron accelerators: (i) the linear accelerator (linac), (ii) the booster and (iii) the storage ring. The electrons are produced and initially accelerated in the linac. The electrons from the linac are injected into the booster where they are further accelerated before being kicked into the storage ring. However, some synchrotron light sources inject the electrons directly from the linac into the storage ring which is a circularly shaped accelerator where the kinetic energy of the electrons is kept constant and the desired electromagnetic radiation is generated. The readers are directed to two recent books in order to Q3 learn more about synchrotron light sources.[8–10] Figure 2shows some of the most important components in a synchrotron facility from the storage ring of the sample to be studied. The design of beamlines varies tremendously depending on the methods (and wavelength range) to be used. For this reason only a very general overview of a beamline set-up is given in Figure 1. From the user’s point of view, everything starts at the photon source where the synchrotron radiation is produced by moving the electrons in the appropriate way within the magnetic device (see Figure 1). The SR produced by the source is pre-conditioned in the front-end before leaving the tunnel to enter into the optic hutch. It must be highlighted that both the energy range and the brightness of the SR strongly depend on the photon source but their discussion is out of the scope of this paper. The unmonochromatized beam is called a white beam and must be ‘conditioned’ before it can be used to analyse the sample in the experimental hutch. It is not possible to review all possible optics devices but we can highlight just a few: (i) Mirrors to transport the beam (and sometime to focus it as well as to collimate and reject unwanted radiation); (ii) Monochromators to select a particular wavelength (or set of wavelengths) from the incoming beam; (iii) Focusing optic elements (if needed) to match the size of the beam to the requirements of the employed technique and sample; (iv) Slits, filters, diagnosis elements, etc. Monochromators can select a very narrow energy bandwidth (usually E/E=10 −4or smaller) for instance using a double-crystal-monochromator and this radiation is called monochromatic, but it can also select a 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 6M.A.G. Aranda Colour online, B/W in print Figure 2. Top left: A diagram of a typical storage ring with the main components depicted, including the circular vacuum chamber (where the electrons are confined), the injector magnet (the system that allows to inject new bunches of electrons), a radiofrequency cavity (that addsenergy to the electrons to keep their kinetic energy constant) and three different typesof magnetic devices that produce the required electromagnetic radiation. Right: The three typesof magnetic devices to produce the electromagnetic radiation: the curved bending magnet (top), and the straight section systems: wiggler (intermediate) and undulator (bottom). Bottom left: Simplified scheme with the components of a beamline: monochromator, focusing optics, sample environment and detector system (a DAC is shown as an example of ancillary equipment). much wider set of wavelengths (f.i. E/E = 1–2% with a multilayer monochromator) for having Q4 higher flux and this type of radiation is commonly named pink beam. The conditioned SR is then used in the experimental hutch to analyse the sample. The appropriate sample answer is recorded in the detector that for combined studies can be more than one. Most of the synchrotron BLs are in the hard X-ray domain (roughly between 4 keV and 50 keV) or soft X-ray domain (roughly between 200 eV and 4000 eV). The optics (and sample set-ups) of soft and hard X-ray BLs are quite different as low-energy implies vacuum which places a lot of restraints in many experiments. It must also be mentioned that IR spectro-microscopy BLs are also starting to play a role in cement studies. The sample (as well as some optics elements) must be positioned with high mechanical accuracy. Hence, high resolution mechanical and piezoelectric stages are commonly used. Custom-made sample environments also allow in-situ experiments. This term should be understood to include, but not be limited to, time-resolved works (under the influence of external parameters like chemical gradients (water, CO2, etc.), temperature, pressure, etc.; or combinations!) usually in the subsecond to hour timescale which is the relevant scale for building material experiments. The main properties of SR that benefit building material studies are summarized in Table 1. The outstanding properties of SR have encouraged their uses in many researches related to cements. Comprehensive information regarding the uses of synchrotron can be found in recent books.[7–10] Very useful properties of SR like its intrinsic time structure (that allows following 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 Crystallography Reviews 7 Table 1. Main properties of synchrotron radiation of interest for building material studies. Property of the SR Features / uses Extreme brightness and flux (combined with high collimation) •Very high signal-to-noise ratio, allowing high quality data •Fast data collection, including kinetics studies (typ. >1ms) •Complex sample environments that absorb radiation, for instance in high temperature, high pressure studies, etc. Radiation damage Energy tunability (from infrared to very high-energy X-rays) •Chemical sensitivity through the absorption edges •High-energy minimizes absorption in complex sample environment experiments •High energy allows to access to large momentum transfer values •Enhance contrast in soft X-ray imaging, f.i. ‘water-window’ Beam size tunability (from tens of mm to tens of nm) •Enable a very large range of imaging techniques from cm in paleontology to nm in nanotechnology •Matches the size of the beam to the experiment requirements •Allows to design sample environments with large degree of freedom Parallel geometry •Improved full-field reconstruction for imaging •Minimize the errors in powder diffraction Limited magnification of X-ray beams Partial transverse coherence •Used in phase contrast imaging techniques •Allows coherent diffraction imaging Stability •High resolution and high reproducibility Notes: Characteristics in italics note possible drawbacks. Polarization properties and temporal resolution due to the electron bunches are not included as they have not been used in cement studies. ultra-fast processes in the picosecond to nanosecond range) and polarization features (that allows studying magnetic and chiral samples) are not highlighted in Table 1as, to the best of my knowledge, they have not yet been used in cement characterization. Also to the best of my knowledge there is no general review about the uses of synchrotron radiation for characterizing cement binders. I am aware of three focused reviews and they are discussed at the beginning of the next section. However, there are general reviews about the uses of synchrotron radiation applied to other fields like cultural heritage.[11,12] In particular, I found the second, very general, work [12] extremely enlightening. Furthermore, some techniques not treated here (because they have not been applied to building materials) are gathered and discussed there. It must also be noted that the evolution of synchrotron radiation, including its growing importance in crystallography, has also been reviewed [13] and also the starting uses of XFEL.[14] In this review article, I highlight recent uses of SR applied to a better/deeper characterization of building materials through selected examples. I restrict the discussion to the last decade and references to previous works are only carried out for a few selected cases. I convey my apologies Q5 to these authors whose work(s) are not referenced here but it would not be possible to account for all works dealing with SR and building materials. I have selected examples to show the wide range of applicabilities, and this work is not intended to be a thorough work compiling all references but the updated summary of some key developments. 3. Uses of synchrotron radiation for cements, pastes and concretes There are three recent reviews devoted to the characterization of building materials by synchrotron characterization techniques. Prof. Monteiro’s group has reported two reviews [15,16] 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 8M.A.G. Aranda Figure 3. Summary of the different synchrotron techniques employed for characterizing cements and pastes. The interaction of the SR can be elastic where momentum is transferred (yielding the scattering techniques) or inelastic where energy is exchanged (resulting in the spectroscopic techniques). Furthermore, there are many imaging tools profiting from these interactions as well as others like absorption. Moreover, time-resolved studiesallow to study kinetics of chemical reactions which are key to follow the evolution in pastes in different time domains. (Acronyms have been defined in the beginning of this paper.) mainly focused on their own activities covering some techniques: (i) full-field soft X-ray nanotomography; (ii) scanning transmission X-ray microscopy; (iii) scanning X-ray microdiffraction microscopy; and (iv) high pressure X-ray powder diffraction. On the other hand, Prof. Provis’s group has also reported one review,[17] mainly focused on their own activities arising from BL-based studies (synchrotron and neutron), dealing with the multiscale understanding of the microstructure and chemistry of geopolymer binders. In this case, the reviewed synchrotron techniques were: (i) scanning infrared spectro-microscopy; (ii) scanning fluorescence X-ray microscopy; and (iii) X-ray nanoand micro-tomographies. With the information that I have gathered, Figure 3displays a scheme containing the different SR techniques that have been employed for the characterization of cements and pastes. Figure 4schematically represents the set-ups used for the main characterization techniques discussed here. As it can be seen from both figures, many synchrotron techniques have been used for the characterization of building materials. These techniques range from scattering for the analysis of phases as well as phase evolutions to spectroscopic tools for a better characterization of chemical species and elemental compositions. Spatially resolved information can be obtained from a number of approaches. In this context, it is appropriate to distinguish between mapping (or raster-scanning) which consists in the sequential measurement of data from adjacent regions of a sample achieved by moving each region into the photon beam, and (full-field) imaging where an image of the sample is recorded onto an array detector where the signal coming from each region of the sample is measured at each pixel. In addition to the mapping and imaging approaches, hybrid approaches are also being developed. There is no doubt that spatially resolved data are very important in building material characterization as they are being used to reveal the complex hierarchical microstructure of the hydrated pastes without alteration, which is not the case for other characterization techniques like electron microscopies. 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 Crystallography Reviews 9 Figure 4. Schematic representation of the main experimental methods used to study cements and discussed in this paper. Reprinted with modification from reference 10 with permission from Elsevier. Figure 5. Main X-ray focusing optics used at synchrotron facilities. KB: Kirkpatrick–Baez mirrors, FZP: Fresnel zone plates, CRL: Compound refractive lenses, Mono and poly-capillaries. KB mirrors and capillary optics are achromatic reflective optics therefore more suited to focus X-rays on wide energy ranges. Reprinted from reference 10 with permission from Elsevier. For mapping and combined approaches, the synchrotron beam must be focused on a tiny spot of the appropriate size (see Figure 4). However, focusing the synchrotron X-rays is challenging since the optical index nof most materials is close to unity for these energies. This is being circumvented by several approaches and Figure 5gives the four main types of focusing optics. For a deeper insight, the reader is directed to a recent review [18] where the progress in the development of hard X-ray microscopy techniques for material characterization at the nanoscale was discussed. Reflective optics (KB mirrors and capillaries) are achromatic, allowing stable beam focusing even when the energy of the incoming X-ray beam varies. Therefore, they are suited to X-ray 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 16 M.A.G. Aranda aluminate reactivity. Therefore, I highlight the work studying the formation and reactivity of five calcium sulphates studied by SXRPD and the Rietveld method.[77] More complex phases can be studied in-situ, both formation and decomposition. As an example: a slurry of synthetic ettringite was heated to form AFm-14 (and bassanite), at about 115°C, on saturated water pressure. On further heating, AFm-14 decomposes to yield hydrogartnet above 210°C. A full transformation pathway was reported by time-resolved SXRPD using a capillary cell.[78] The thermal decomposition of other cement phases has also been studied by SXRPD, like 3Mg(OH)2·MgCl2·8H2O, the main phase of Sorel cement.[79] High-temperature SXRPD has been used to follow the thermal transformations of special cements or mixtures. This is the case of cement-asbestos decomposition studies.[80,81]The highest reported temperature was 900°C as the authors used an air blow heating system with the sample within quartz capillaries. This type of preliminary studies is important for safe treatments of asbestos in novel industrial reactors. 3.4.2. In-situ clinkering studies at very high temperatures High-temperature SXRPD can also be used for the in-situ study of the reactions taking place in the clinkering process. These are complex experiments as the temperature can be as high as 1400–1500°C with important fraction of the sample melt, and so the capillary must be inert towards this very reactive environment. Using a halogen furnace that can heat up to 1600°C, the clinkering reactions of active belite clinkers were reported using platinum tubes.[82]The selected wavelength was very short, λ= 0.30Å, to go through the highly absorbing Pt capillary. Rietveld methodology was successfully applied to the recorded data to obtain quantitative phase analyses. In a subsequent work, the same experimental conditions were used to study the clinkering reactions of iron-rich belite sulphoaluminate cements, aka sulphobelite.[83] In this study, the high-temperature reactions were established and quantified and the role of borax for activating these cements was also characterized. Figure 8shows a sequence of steps to carry out this type of in-situ experiments. First, the precalcined raw material mixtures(s) are loaded into Platinum tubes.[82] A precalcining step is needed to release the CO2and avoid overpressures in this type of experiments, where the sample is sealed within the capillary. Furthermore, expensive Pt capillaries/tubes are needed because the heated mixtures contain a relatively large fraction of highly reacting liquid (above 1200–1400°C), and so inexpensive ceramic tubes/capillaries cannot be used. The Pt tube is isolated from the diffractometer by anMgO ceramic bar which is bonded by a refractory glue (dark powder between Pt and MgO in Figure 8, top left). The cell is assembled, mounted within the diffractometer axis and the temperature is coarsely controlled by the calibrated voltage applied to the halogen lamps. Then, the SXRPD patterns are collected at (approximate) temperatures. The temperature can be accurately known from the values of the refined unit cell parameters of platinum, as its thermal expansion is well known. Finally, RQPA is carried out based on the pattern regions which contain diffraction peaks from the sample and are not severely overlapped with the Pt diffraction peaks. Under thesevery tough conditions (high temperatures with samples containing about 30 wt% of a highly reactive ceramic melt), the SXRPD patterns were of sufficient quality to allow appropriate quantification of the phase evolution.[82,83] 3.5. Diffraction under pressure to characterize phases and chemical reactions In this section, I will review some SXRPD studies of single phases or cements under pressure. I have subdivided this part into three sections depending upon the level of pressure reached in the different approaches. Hydrothermal reactions take place at relatively low pressure, usually 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 Crystallography Reviews 17 Colour online, B/W in print Figure 8. (Top left) Photography of the disassembled halogen lamp cell that allows in-situ heating up to 1600°C. Two Pt tubes, loaded with the raw materials, and isolated from the goniometer head mounting system by a MgO refractory, white ceramic, are also shown. (Top right) Photography of the assembled cell mounted in the ID31 synchrotron powder diffractometer (ESRF). (Bottom right) Raw SXRPD data collected at high temperature where the diffraction peaks of Pt holder (tube) dominate the scattering within the pattern. (Bottom left) Rietveld quantitative plot of the appropriate region of the previous pattern; where the diffraction peaks from the clinker phases are present. lower than 20 bars (2 MPa). There are capillary cells that can withstand up to 200 bars (20 MPa) and these works are reviewed in the second section. Finally, aDAC can withstand more than 1 Mbar (100 GPa) and is normally used to study the high pressure behaviour of hydrated cement phases to derive mechanical properties like the bulk modulus. These works are reviewed in the last subsection. 3.5.1. Chemical reactions at relatively low pressures in capillary cells (hydrothermal conditions) The formation reactions of AFt and two types of AFm phases were studied by SXRPD and hydrating different calcium aluminate phases in the presence of gypsum.[84] The reactions were investigated with the samples loaded in quartz capillaries, and the maximum internal pressure was 17 bars, which was obtained by supplying N2gas. The studied temperature range was 25– 170°C, which was obtained by heating the capillary with a hot air flow system. In a second work,[85] these authors studied the hydrothermal transformations of the calcium aluminium oxide hydrates, CaAl2O4.10H2O and Ca2Al2O5.8H2O, to yield crystalline Ca3Al2(OH)12, and 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 18 M.A.G. Aranda where intermediate phases were identified. On the other hand, instead,a capillary cell, an autoclave cell with Be windows was designed and built,allowing the study of cements by SXRPD up to 12 bars and 200°C.[86] This autoclave cell was initially used for studying the hydrothermal formation of tobermorite under different conditions [86,87] and later it was used for investigating the phase evolutions during the production of autoclaved aerated concretes.[88–90] 3.5.2. Hydrating reactions at medium pressures and temperatures Several hydrothermal reactions take place at pressures up to 20 bars, but there are other conditions where (much) higher pressures apply. For instance, deep oil-well cement slurries can be subject to pressures of up to 1 kbar. Therefore, to study cement hydration in the range of 100– 200 bars it is very interesting for several applications including oil-well cements which surround metal oil-well liners to form a gas tight seal between the bore wall and the liner. For this type of studies, a cell was designed and built for SXRPD data collection, in optimum conditions, Q7 up to 500 bars and 200°C.[91] Later, this cell, which performs best with sapphire capillaries, has been used for studying a number of hydrating cement slurries including: (i) Class A and H oil well cements mixed with variable amounts of CaCl2for accelerating the hydration [92]; (ii) Class H oil well cements mixed with silica flour, silica fume and a natural zeolite which are used to improve the mechanical properties of the binder [93];(iii) White cement, class G and class H oil well cements, without additives, for analysing the reaction kinetics of the main phases [94];and (iv) C3S in the presence of several retarders to counterbalance the accelerating behaviour of pressure.[95] These authors have also developed a cell to allow simultaneous measurement of SXRPD data and ultrasound shear-wave reflection data from cement slurries at variable temperatures and pressures.[96] 3.5.3. Phase characterization at very high pressures in diamond-anvil-cells The pressures employed in this type of studies (dozens of GPa equivalent to hundreds of kbars) are much higher than what could exist in cement pastes (mortars or concretes);however these works are important to derive mechanical properties of the hydrated phases as well as to study the stability of cement phases which may help to understand reaction mechanisms. The pressure transmitting medium is important as it should not induce any structural modification to the studied phase and it must be easy to work with. Silicone oil as well as methanol/ethanol mixtures have been used for studying hydrated cement phases at high pressures in DAC systems. Figure 9shows a scheme with components of a DAC. There are different types of DACs, but its discussion is out of the scope of the present work. When the sample is loaded within the DAC, the set-up is placed in the diffractometer and the micro-focused beam interacts with the sample. In order to reach higher pressures, smaller samples are employed. The diffracted beam is recorded in an appropriate 2D detector (commonly a CCD detector). Forces are varied with the pressures being measured, commonly, from the fluorescence signal of a ruby chip. Finally (from the data collection point of view), the 2D patterns are radially integrated to transform them to 1D patterns (see Figure 9) that can be analysed with any Rietveld package. The DAC approach has been used to study the behaviour of ettringite up to 6 GPa. The two main outputs of this study were the isothermal bulk modulus of ettringite, 27 GPa, and its transformation to an amorphous phase at pressures above 3 GPa.[97] The bulk modulus of gel C-S-H is of utmost importance as it is the main binding component in OPC mortars and concretes. However, as its chemical composition can vary, to report a single number is complicate. In an initial work, two samples were studied, synthetic C–S–H(I) and another gel obtained from the hydration of alkali-activated slag. The powder diffraction patterns were collected up to 4GPa and the bulk modulus for the two gels were very similar, ≈35 GPa, showing that Al substitution in 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 Crystallography Reviews 19 Colour online, B/W in print Figure 9. Schematic representation showing the experimental work flow in a high-pressure experiment using DAC. (Top right) Scheme of a diamond-anvil-cell with its different components including the ruby chips commonly used to measure the applied pressure. The (micro)focused SR X-ray beam is diffracted by the sample and the high-pressure pattern is collected in a suitable 2D detector. (Top left) A CCD 2-D pattern is collected at a given pressure. Then, the 2D pattern is integrated to yield the typical intensity versus angle 1D pattern. (Bottom) Finally, several data sets are collected at different pressures and the appropriate information is derived (f.i., unit cell variation, crystal structures, amorphization processes, bulk modulus, etc.). C–S–H(I) does not significantly affect the bulk modulus of C–S–H(I).[98] On the other hand, the two crystalline phases more similar to the ill-crystalline C–S–H gel are 14Å-tobermorite and jennite. Hence, SXRPD data for 14Å-tobermorite were measured up to 5GPa and the obtained bulk modulus was 47GPa.[99] A similar study for jennite, measured up to 6 GPa, yielded a bulk modulus of 64 GPa.[100] The isothermal bulk modulus of aluminium-substituted 11Å-tobermorite from relict lime clasts of 2000-year-old Roman seawater harbour concrete was also measured, yielding Ko=55GPa[101]. Other works reported the bulk modulus for selected samples of interest in cement chemistry like stratlingite: Ko= 23 GPa, hemicarboaluminate: Ko= 15 GPa, monocarboaluminate: Ko= 54 GPa and hydrogarnet: Ko= 70 GPa.[102,103] X-ray diffraction is the standard method for measuring the unit cell volume variation with the pressure of crystalline materials. For a pressure range where the composition of the phase is constant, this directly translates into the density variation with pressure. Using diffraction to determine the density variation of non-crystalline materials is not straightforward, so alternative methods are being investigated. Recent advances have allowed XAS data, taken for an amorphous material loaded in a DAC, to be used to characterize the density variation with pressure. This approach has been successfully used in cement chemistry.[104] The elastic properties of an ASR gel were studied by XAS and Brillouin spectroscopy measurements. XAS was used to determine the density of the gel as a function of pressure, yielding an isothermal bulk modulus 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 20 M.A.G. Aranda of 33 GPa. Brillouin spectroscopy was applied to measure isentropic bulk (24.9–34.0 GPa) and shear moduli (8.7–10.1 GPa) of the gel. Finally, SXRPD data collected in DAC as a function of pressure have also been recorded for anhydrous phases of interest in cements. Experimental data were recorded for C3A which yielded an isothermal bulk modulus of 110 GPa which was compared to the results from theoretical calculations.[105] Other elastic properties were also measured and compared to the theoretical results, a second example being ye’elimite. The reported isothermal bulk modulus was 69GPa.[106] 3.6. Small angle X-ray scattering for microstructural characterization of building materials When anX-ray (or a neutron) beam passes through a material, under the appropriate experimental conditions, a component of the direct beam is scattered out at a small angle due to heterogeneities (microstructure) of the sample. The angular profile of the SAXS (or SANS) intensity is effectively a Fourier transform of this microstructure. The recorded data are commonly analysed using appropriate microstructure models, and the microstructures are described/quantified through the resulting parameters. Although SAXS data can be obtained with laboratory sources, the quality of the data from SR instruments is much higher. Therefore, most of the SAS studies are carried out at synchrotron (or neutron) facilities. A focused review was devoted to the applications of SAXS and SANS characterization techniques to C–S–H gel and cement pastes.[107]Inthis review, the similarities and differences between SAXS and SANS were also highlighted. SAS (SAXS and SANS) techniques are nondestructive tools for characterizing density fluctuations over a wide range of length scales without altering the sample (sample preparation that may modify the microstructure is not commonly needed). Furthermore, it can probe interfaces of closed as well as open pores, unlike mercury porosimetry where information about the closed pores is not available. SAS provides statistically averaged information over the bulk of a material, unlike microscopy techniques where the information is limited to the local studied part of the sample. However, data treatment is important to extract the appropriate information and data analysis is far from obvious. There are several microstructure models ranging from Guinier approximation to fractal morphologies and full analysis of coherently ordered microstructures.[107,108] Some examples of uses can be highlighted. By combining SAXS and SANS data with several other techniques, the mean formula and mass density of the nanoscale C–S–H gel particles in untreated hydrating cement were determined to be (CaO)1.7SiO2.1.80H2O and 2.604g/cc, respectively.[109] These values differed from previous reported ones for the C–S–H gel, likely because of the specific drying conditions used in different approaches. Other study focused on determining the pore volume fraction of hydrated cement compacts from SAXS data and their evolution with hydration time.[110] The changes in the microstructures due to the use of cement additives have also been investigated by SAXS and SANS (see Figure 10).[111] The reported results showed that the used PCE tends to increase the size of the disk-like C-S-H globules but haslittle influence on the thickness of the water and calcium silicate layers within the globules. It must also be noted in importance of SANS studies. As a key example, this technique has been Q8 used for monitoring the degree of homogeneity of a cement paste matrix with the final goal of immobilizing lowand intermediate-level radioactive waste.[112] Finally, SAXS data have also very recently been used to validate 2D high-resolution observations by TEM. Under some approximations, and using the Fourier slice theorem, it has been shown that the spectral density from TEM is a good approximation of the SAXS pattern. This have been used for a deeper characterization of the C–S–H gel where the computed SAXS signal (from TEM) and the measured spectrum agree quite well within the limited overlapping q–range (0.01to0.04Å −1).[113] 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 Crystallography Reviews 21 Colour online, B/W in print (a) (b) (a) (b) Figure 10. (Left panel) Model fitting results of a pure C–S–H sample. (a) Experimental data of SAXS (blue open circle) and SANS (black open circle) and the corresponding data fitting curves of SAXS (cyan line) and SANS (red line). (b) Inter-particle structure factor S(Q) of the SAXS data (blue solid triangle) and SANS data (black solid circle) and intra-particle structure factor P(Q) of the SAXS data (green open triangle) and SANS data (red open circle) used to fit the data in panel (a). (Right panel) Model fitting results for (a) SAXS data and (b) SANS data. Both panels show the experimental data for pure C–S–H (black open square), and C-S-H with different amounts of superplasticizer. For further details, the reader is directed to the original publication. Reprinted from reference 109 with permission from Elsevier. 3.7. Imaging building materials at different length scales and with different photon energies There are many different types of imaging methods based on different wavelengths (IR, soft X-rays and hard X-rays) as well as on different experimental set-ups. Figure 11 displays a schematic representation of the three more common techniques: full-field microscopy, scanning transmission microscopy and coherent diffraction imaging. All these three set of techniques (with subgroups) have been employed for characterizing the rich hierarchical microstructure details of cement pastes. A recent publication has reviewed state-of-the art X-ray imaging techniques based on partially coherent synchrotron radiation including: full-field tomography, scanning transmission microscopy, ptychographic forward coherent diffraction imaging and scanning small-angle X-ray scattering [114] but it did not specifically deal with cement research. Another recent review [18] discussed the progress in the development of hard X-ray microscopy techniques for materials’characterization at the nanoscale. They reviewed state-of-the-art hard X-ray synchrotron nanoscale microscopy techniques which included 3D tomographic visualization, spectroscopic elemental and chemical mapping, microdiffraction-based structural analysis, and coherent methods for nanomaterial imaging. For full-field imaging, see Figure 11 (top);the synchrotron beam is relatively large and the ability to resolve tiny details of the sample arises from either a magnified projection onto the X-ray detector or from amagnifying lens. In the former approach, image resolution is limited by the pixel size of the detector and by the size of the X-ray source; in the second case, resolution 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 22 M.A.G. Aranda Colour online, B/W in print Figure 11. Scheme of the three main techniques for X-ray imaging highlighting some limitations. (Top) full-field microscopy, where the sample is fully bathed by the X-ray beam. (Intermediate) scanning transmission microscopy, where the sample is raster scanned by a focused beam. (Bottom) Coherent diffraction imaging, where the image is generated by an appropriate reconstruction algorithm. In the three cases, proper sample rotation allows recording tomographic data. and image quality depend on the X-ray optics. The challenge to increase the resolution resides in producing better X-ray optics elements that combine high numerical aperture, high efficiency and low aberrations. This is assuming that radiation sample damage does not limit the achievable resolution. For scanning imaging, see Figure 11 (intermediate);the beam is focused on a tiny spot that it is raster scanned to deliver 2D (radiographic) or 3D (tomographic) images. Furthermore, different properties can be measured (absorption, fluorescence, diffraction, etc.) which provide complementary information. For instance, STXM is areal-space imaging technique that utilizes focusing optics (it could be a Fresnel zone plate or a capillary) to deliver a small monochromatic X-ray beam onto a sample and it measures the intensity of the transmitted beam in raster mode,thereby filling an image array. STXM does not require an objective lens, and hence there is no attenuation of the beam between the sample and the detector, which means that the dose delivered to the sample is reduced compared to a full-field, lens-based approach. The procedure allows imaging with classical spectroscopic techniques;see below, at sample-tailored X-ray photon energies. There is an alternative image-forming approach where the X-rays scattered by the sample are analysed and reconstructed by appropriate mathematical algorithms, see Figure 11 (bottom). Such lensless techniques are often grouped under the term: CDI or diffraction microscopy (not to be confused with X-ray diffraction micro-tomography, see below). It is widely believed that the CDI technique allows image resolution limited only by the tolerance of the samples to radiation damage. Furthermore, single-shot experiments at ultra-fast XFEL sources may even bypass this limitation. In a typical CDI experiment, a small sample is fully bathed with a plane wave (fullfield technique), and its far-field scattering pattern is recorded in oversampling conditions. For more information about CDI, the interested reader is directed to a recent review.[115]Atthis stage, two related techniques should be mentioned that depend on where the detector is placed. In BCDI, the far-field scattering pattern is collected at the right angle of a diffraction peak for a crystalline microparticle. In FCDI, the far-field scattering pattern is collected at the small angle scattering and the sample does not need to be crystalline. While the scattering process is well known and understood, inverting measurable intensity distributions to obtain an image of the 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 Crystallography Reviews 23 sample, often known as image retrieval, is a mathematically ill-posed problem where a unique solution has to be ensured. Image retrieval usually comprises many cycles of alternate iterative processes enforcing aprioriknowledge on the sample, such as its finite size and consistency with measured data. Such a‘classical’ CDI approach requires the samples to be isolated and has very low tolerance to perturbations by signals due to other scatters, f.i. from ice. Very importantly, there is an evolution of CDI where the samples are scanned by a small beam overlapping part of the illuminated samples (ptychographic approach) which favour/ensure the convergence of the reconstructing algorithms. 3.7.1. Full-field soft X-ray nano-tomography Transmission soft X-ray full field microscopy was originally developed to study biological samples, but later has been applied to the study of the hydration processes of cement-based materials. Working with soft X-rays of an energy of ≈520 (eV), which corresponds to a wavelength of 2.4nm, it enhances the contrast between the cement particles and developing hydrates which contains more oxygens ‘water window’. Soft X-ray microscopy allows high-resolution imaging of hydrated samples over time and permits complete imaging of samples up to about 5 µm thick. For building materials this limitation effectively means loading dilute suspensions with very high w/c ratios (a severe drawback for many types of studies). In order to prevent excessive dissolution of the cement grains, several works used saturated solutions of CH and/or CSH2. However, saturated solutions cannot overcome the other artefact of dilute solutions, the increased availability of space in which hydration products are growing. Initially 2D soft X-ray full-field microscopy, ACsX-nM, was used to directly imagethe hydration of C3S and OPC in a saturated solution of CH and CSH2with time.[116] Figure 12(a) shows aC 3S grain which has needle-like hydration products on its surface likely showing the of C-S-H gel formation after the induction period. Figure 12(b) and 12(c) showsthe hydration evolution with time of OPC which clearly shows the formation of hexagonal prisms ettringite crystals. ACsX-nM has also been employed to study the differences in real-time hydration between pure (cubic) tricalcium aluminate and Na-doped C3A (orthorhombic) in aqueous solutions saturated in sulphates. The reactions involving cubic C3A were more influenced by higher concentrations of sulphate ions, forming smaller ettringite needles at a slower pace than for orthorhombic C3A. It was also concluded that the rate of release of aluminate species into the solution was also accelerated by Na-doping.[117] The early-age hydration of ye’elimite phase (the main component of the calcium sulphoaluminate cements) has also been studied using this technique.[118] Diluted suspensions of this phase saturated in CSH2, and with variable amounts of CH, were examined. The most voluminous hydration product observed was ettringite. Under these hydration conditions, AFt commonly displayed acicular, filiform and plenty of intergrowths with stellate habits (see Figure 13). The previous examples used 2D image evolution (microscopy) for obtaining the required information. The current technology allows to rapidly record a set of 2D images rotating the sample and so tomograms can be acquired and reconstructed. Therefore, as images may have nanometer resolution, the term nano-tomography was coined. ACsX-nCT has been recently employed to characterize the microstructure of two hydrated cement pastes.[119] Furthermore, it was possible to perform a direct computation of the ultra-small angle-scattering spectra from a single projection image which agreed well with the experimental synchrotron SAXS data obtained for the same pastes. Finally, soft X-ray nanotomography has also been employed for characterizing 2000-year-old Roman seawater concrete.[101] The data analysis showed clusters of Al–tobermorite crystals displaying both platy and elongated 1–2µm crystals typical of geological 11 Å–tobermorite. 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 24 M.A.G. Aranda (a) (b) (c) Figure 12. In-situ soft X-ray images, ACsX-nM, for (a) C3S hydrating for 8h 51 min in a diluted solution saturated with CH-CSH2; (b) OPC sample hydrating for 15 min in a diluted solution saturated with CH-CSH2; (c) as (b) but hydrating for 55min. Scale bars correspond to 1 µm. Reprinted from reference 114 with kind permission from Springer Science and Business Media. 3.7.2. Full-field hard X-ray micro-tomography Full-field transmission hard X-ray computed micro-tomography (here summarized as AChXµCT) is the high-resolution adaptation of the CAT technique employed in medical applications since the 1970s. The AC-hX-µCT imaging technique is based on the virtual reconstruction of the inner density distribution of a sample from 2D X-ray radiographs collected at many viewing angles, while the sample rotates relatively to the source-detector direction (see Figure 14).We can distinguish two different steps in data handling. Initially, the recorded data must be processed to obtain a digitalized version of the studied object. The final outcome of the reconstruction process is a series of images called slices that, when stacked together after proper alignment,provide a 3D map of the spatial variations of the X-ray linear attenuation coefficient (µ) within the investigated object. Here, the key result is that each slice is composed by a matrix of voxels (volume elements) whose grey values are proportional to the mean value of µ(see Figure 14, bottom) The maximum spatial resolution achievable can reach the sub-micrometre scale for standard applications and is related not only to the technical specifications of the experimental set-up (characteristics of the synchrotron X-ray source and used detector), but also and chiefly to the size of the sample. As a general rule, higher resolutions require smaller samples due to the limited field of view of current X-ray area detectors. In a second stage, the digitalized object must be analysed with the appropriate software(s) depending onthe information to be obtained: pore size distribution and connectivity, tortuosity, size and shapes of some particles, leaching effects, etc. The main advantage of AC-hX-µCT is that it provides three-dimensional visualization of the internal microstructure of untreated samples. This is key as several other techniques for microstructure analysis (MIP, SEM, TEM) are known to produce irreversible changes in the 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 Crystallography Reviews 25 Figure 13. In-situ soft X-ray images, ACsX-nM, of hydrating C4A3S particles in a saturated CH-CSH2 very diluted solution, showing the ettringite growth. Hydration time is indicated. Scale bars correspond to 1µm. Reprinted from reference 116 with permission from Elsevier. pore structure of cement pastes, particularly at small sizes. Furthermore, parameters such as connectivity and tortuosity are completely inaccessible by these techniques but they can be inferred from AC-hX-µCT, although the obtained results can be resolution dependent, see below. Fast data acquisition times are very important to avoid heating of the sample with the possibility to alter the microstructure of the studied sample due to local dehydration processes. The main two disadvantages of AC-hX-µCT are: first, its still low spatial resolution, voxel size of ≈0.5 µm3, relative to the sizes of capillary pores controlling transport properties in mature pastes. Second, sample preparation is delicate as narrow capillaries are required to acheive high resolution and in these conditions the w/c ratio and homogeneity of the pastes are difficult to ensure. Perhaps it is worth mentioning that initial (independent) contributions from the Europeans, Japanese and Americans were reported within two years. In a seminal work of European researches,[120] AC-hX-µCT was used to study the connectivity and tortuosity of the pore network in OPC pastes. It was also shown that the degree of connectivity of the pore network was very sensitive to both the spatial resolution of the images and the evolution of contrast resolution during ageing of the cement. Some experimental conditions were: Lindemann Glass type capillaries with diameter of 600 µm and a wall thickness of 10 µm; pastes with w/c ratio of 0.5; X-ray 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 32 M.A.G. Aranda strätlingite was observed in the system containing 80% of fly ash, confirming that the amount of alumina and silicate phases provided by the fly ash is a major factor for the formation of stratlingite (and C–S–H). One selected example on the combined use of hard X-ray SFXM and micro-EXAFS and micro-XANES spectroscopies is the work on the Ni uptake by OPC pastes including the influence of the inherent heterogeneity of the cement matrix on the Ni speciation.[145] Some key experimental details were: beam size of 5 ×5µm, a fixed beam energy of 10keV for the µXRF study and a variable beam energy (close to the Ni K-edge, 8.3keV) for the µ-XAS study. Both µ-XRF and µ-XAS data were collected at room temperature in fluorescence mode using a seven-element Ge-solid-state detector. Unfortunately, the thicknesses of the thin sections were not reported. 3.7.3.4. Scanning diffraction hard X-ray microscopy In section 3.3, the uses of synchrotron powder diffraction for quantifying crystalline phases and to follow phase evolutions are discussed. These works used a large beam (usually larger than a millimetre) yielding an accurate average picture but without spatial resolution. However, for heterogeneous materials, like cement binders, and for some applications, it is invaluable to have the spatial distribution of the different phases (for instance the changes/alteration with depth due to sulphate attack). This can be obtained by focusing the beam down to several micrometresize and scanning the sample in the appropriate direction. The technique, SDXM, works in transmission but not only is beam focusing important (for instance using K-B mirrors) but sample preparation is also very important. For this technique to be useful, a relatively thin (unaltered) cross-section must be prepared along the appropriate direction (see Figure 19). Typical thicknesses for the flat slices used in these studies ranged from 100 to 500 µm. This technique is commonly known simply as synchrotron microdiffraction. SDXM, aka synchrotron microdiffraction, was used to quantify the orientation distribution of fibrous ettringite crystals after sulphate attack in fractured concretes.[146] Sample preparation is key and a summary of the process follows. A flat surface was impregnated with epoxy resin and then mounted on a glass slide. Then, using a diamond saw, and with kerosene as cooling agent, a 50 µm thin slice was prepared also sealed with epoxy. Finally, the slice was removed from the glass slide and used for the synchrotron characterization (beam energy: 8keV, beam Colour online, B/W in print Figure 19. Schematic representation of the sample preparation and experimental method for scanning hard X-ray diffraction microscopy, aka synchrotron microdiffraction. Reprinted with permission from reference 145. Copyright {2011}American Chemical Society. 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 Crystallography Reviews 33 size: 2 ×2 µm). The analysis revealed that the c axes of the trigonal ettringite crystallites were preferentially oriented perpendicular to the fracture surfaces. SDXM was also employed to determine, with spatial resolution, the changes in the phase composition due to sulphate attack [147] for establishing the durability of cementitious materials under these conditions. Key experimental details were, beam energy: 11.6 keV, beam size: 10 µm, sample thickness: 200 µm. Furthermore, the sulphate attacks are affected by the presence of SCM. In two subsequent works from the same group, the microstructural profile analyses of concrete deterioration after sulphate attack of OPC blended with fly ash [148] and with natural pozzolana, granulated blast furnace slag or fly ash [149] were deeply investigated. In the first work,[148] OPC was mixed with 30 wt% of class F fly ash and hydrated with w/c ratio of 0.5 for 28 days. Afterwards, the sulphate attack was carried out under laboratory conditions for 6 months. Then, the samples were embedded in epoxy resin and polished to thicknesses of 200 µm. During all preparation steps the samples were cooled with petroleum to avoid dissolution of water-soluble phases. Key experimental details were, beam energy: 14.5 keV, beam size: 10 µm. In the second work,[149]thesample specimens were embedded in sulphate-bearing soil (≈1wt%SO2− 4) for 19 years. The samples were also embedded in epoxy resin for the preparation of cross-sections which were made parallel to the direction of the sulphate ingress. Selected experimental details were, beam energy: 12.4 keV, sample thicknesses: 500µm. The beam size was not reported but it was mentioned that the achieved spatial resolution was 30 µm. Finally, synchrotron microdiffraction can be used to map the phase distributions in any complex binders after successful sample preparation. For instance, this technique has been very recently applied to study pyroclastic aggregate concrete of Trajan’s Markets (1900 years old) as well as their reproductions (experimental archaeology): hydrated lime–volcanic ash mortar that binds decimeter-sized tuff and brick aggregates.[150] The mortar reproduction gains fracture toughness over 180 d through progressive coalescence of C–A–S–H gel binder and crystallization of strätlingite and katoite at ≥90 days, after pozzolanic consumption of hydrated lime was complete. Key experimental details were, beam energy: 10 keV, beam size: 8 ×2 µm, and sample thickness: 300 µm. 3.7.3.5. Scanning diffraction hard X-ray micro-tomography SDX-µCT [151] is a combination of diffraction (crystalline phase sensitive) and imaging (through tomographic reconstruction) techniques which allows determining the three-dimensional spatial distribution of different phases within heterogeneous samples. The reconstruction scheme for the SDX-µCT technique is depicted in Figure 20. Unless the previous technique, SXDM, flat thin sections are not needed. Q9 Furthermore, the appropriate rescaling of the voxel intensity to the total intensity of sample scattering makes it possible to obtain the absolute quantification of the phase proportions in each voxel. SDX-µCT was initially used for 3D monitoring of the evolution of the microstructure and phase formation non-invasively.[152–154] For this particular set of experiments, the authors used a monochromatic beam (E= 18keV) with a size of 2 ×4µm2. Although this technique gives good insight into the three-dimensional phase arrangement at intermediate and later hydration ages, the long acquisition times (∼8 h for a slice of 500 µm diameter and 2 µm thickness) do not allow the phase mapping at early stages of hydration, as reaction kinetics are too fast for the microstructure to be resolved. Subsequently, AC-µCT was used at early ages (7 h of hydration) and SDX-µCT at later ages (7 days of hydration) for obtaining better insight into the phase and microstructure developments in hydrating pastes.[155] SDX-µCT, combined with numerical simulations for C–S–H precipitation, was used to investigate the C–S–H topological distribution and modes of precipitation.[156] This technique have been recently used to map the phases present in two hydrating OPC cement pastes (one sample hydrated with pure water and a second one hydrated in the presence of nucleation seeds). The quantitative description of the phase spatial distribution by radial distribution functions 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 34 M.A.G. Aranda Figure 20. Schematic representation of the experimental set-up and workflow (data acquisition and data processing) for scanning diffraction hard X-ray micro-tomography. Reprinted from reference 151 with permission from International Union of Crystallography. allows the discrimination of different nucleation mechanisms.[157] SDX-µCT has also been very recently used to map the C−S−H precipitation in the absence and presence of superplasticizer (see Figure 21,left). The observed spatial correlation between C−S−H and unhydrated cement particle surfaces indicated that, in the absence of PCE superplasticizers, C−S−H forms by a process of heterogeneous nucleation, on the surface of the dissolving cement particles (see Figure 21,right). Conversely, the lack of significant spatial correlation between C−S−H and the surface of unhydrated particles, when PCE is added to the system, revealed that C−S−H precipitates randomly throughout the available space in the paste.[158] 3.7.4. Hard X-ray coherent diffraction imaging As it was mentioned above, there is an alternative to image-forming optics where the Xrays scattered by the sample are reconstructed by appropriate algorithms. These set of techniques are commonly named CDI techniques, and the two most common ones are discussed below. 3.7.4.1. Hard X-ray Bragg coherent diffraction nano-tomography BCDI is also a noninvasive imaging technique which can yield three-dimensional images of individual crystals on the nano-scale through inversion of the diffraction data by acomputational method, but it also highly sensitive to crystal defects and strain fields inside crystals seen as phase evolution. The experimental set-up (see Figure 22)is compatible with cement hydrations as very recently reported.[159] In this study, the early hydration (up to 3 days) of microcrystals of calcium monoaluminate, CA, was investigated in situ by following the 3D Bragg diffraction electron density and strain evolution (see Figure 23). The variation of Bragg density within the crystal was attributed 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 Crystallography Reviews 35 Colour online, B/W in print (a) (b) (c) Figure 21. Scanning diffraction X-ray micro-tomography results. (Left) Phase maps displaying the space distribution of the unhydrated cement particles (red), C −S−H (green) and the combination of the two within a virtual slice through an OPC paste sample hydrating in water and in the presence of a PCE superplasticizer. The colour intensity is proportional to the volume fraction of the given phases in each voxel. The white circle represents the enclosing glass capillary (internal diameter 400 µm). (Right) Radial distribution functions relative to the spatial distribution of C −S−H with respect to the position of the unhydrated particle surfaces, as calculated for: (a) the phase maps of the OPC sample without PCE, at 7 days of hydration; (b) the phase maps of the OPC sample with PCE, at 7 days of hydration; (c) the difference phase map of the OPC sample with PCE at shorter times (7−17 h). Reprinted with permission from reference 156. Copyright {2015}American Chemical Society. to the change of the degree of crystal ordering, which could occur through ion transfer during hydration. The observed strain, coming from interfacial mismatch effect between high Bragg density and low Bragg density parts in the crystal, remained throughout the experiment. The first Bragg density change during hydration process was due to a big loss of Bragg density and was seen as removal of density, but not phase. The work provided new evidence supporting the through-solution reaction mechanism of calcium monoaluminate. 3.7.4.2. Hard X-ray ptychographic forward coherent diffraction nano-tomography PFCDInCT is a non-invasive imaging technique based on the (partly) coherent properties of synchrotron radiation and it allows the three-dimensional mapping of the electron density in the studied sample. This variant of CDI does not require crystalline ordering and its quantitativeness combined with a resolution close to 100nm (for a field of view of about 60–100 µm) makes it very appropriate for studying the hierarchical microstructures in complex materials including cement pastes. The term ‘ptychography’ stems from the Greek word for a fold, related to the Latin origin of the term convolution. Instead of fully illuminating a small sample with a featureless plane wave, 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 36 M.A.G. Aranda Colour online, B/W in print Figure 22. Schematic representation of the experimental set-up for hard X-ray Bragg coherent diffraction nano-tomography. Reprinted with permission from reference 157. Copyright {2015}American Chemical Society. Colour online, B/W in print (a) (b) (c) (d) (e) Figure 23. Cross-sections through the reconstructed 3D images of the CA crystal showing the internal Bragg density variations during 3 days of hydration measured by BCDI. (a) unhydrated CA. (b) CA hydrated for 2h. (c) CA hydrated for 24 h. (d) CA hydrated for 52h. (e) CA hydrated for 67h. Reprinted with permission from reference 157. Copyright {2015}American Chemical Society. 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 Crystallography Reviews 37 Colour online, B/W in print (a) (b) Figure 24. (a) Schematic representation of the experimental set-up for hard X-ray ptychographic forward coherent nano-tomography. At each incidence angle, coherent diffraction patterns are recorded by a pixelated detector for a number of overlapping scanning positions, which allows the projected complex-valued transmission function to be reconstructed. (b) Example of a single reconstructed phase projection of the epoxy resin impregnated hardened cement paste sample. The black dots indicate the scanning positions at which diffraction patterns were recorded, and the black circles represent the approximate shape of the beam – shown for the first two shells of the circular scan only. The scale bar corresponds to 5 µm. Reprinted from reference 158 with permission from Elsevier. it used a small X-ray beam to raster scan an extended sample. The deconvolution of the effects Q10 due to the sample from those due to the structured illumination can be ensured if the sample is scanned in sufficiently fine, overlapping steps (see Figure 24). Ptychography became practical only by combining it with iterative phase retrieval algorithms which reduced the sampling requirements drastically. PFCDI-nCT was applied to image hydrating cement pastes.[160] Figure 24 illustrates the experimental set-up (a) as well as the result (b) for imaging a sample of resin-impregnated, hardened cement paste. The data from multiple known scan positions are inverted to yield a 2D image, whose resolution is limited by the maximum scattering angle where there is signal and by the positioning accuracy of the sample. When combined with a rotation stage/strategy, a 3D tomographic image can be obtained. Furthermore, the high accuracy in measuring the electron density allows accurate segmentation of the data. In a very recent work,[161] PFCDI-nCT has been applied to the microstructural characterization of C−S−H formed by hydrating C3S. The 3D spatial resolution of the phase contrast images was close to 130 nm, whereas the resolution of the absorption images was poorer, ≈250 nm (see Figure 25). It has been observed that the C−S−H density can depend on the particles’ states of hydration. For fully hydrated particles, the estimated density of the outer-product C−S−H was larger than that of the inner-product C−S−H, whereas for the partially hydrated particles, the densities of the apparent outer and the inner products were very similar. The density values of C−S−H ranged from 1.72 gcm−3 to 1.96 gcm−3, and its water content ranges from 4.3 to 7.6 mol, assuming a fixed C/S molar ratio of 1.75. 4. Outlook It is difficult to forecast the main research lines in cements using synchrotron tools as these evolve quite rapidly. In any case, there are challenges in the chemistry of cements where developments in the synchrotron characterization techniques may play a leading role. I highlight my shortlist below: 1. Most cement binders are based on amorphous gels or they contain large contents of amorphous materials. We all know that the characterization of amorphous materials is always complicated due to the lack of long-range order and periodicity, and also because of their 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 38 M.A.G. Aranda Colour online, B/W in print (a) (b) (c) (d) (e) (f) Figure 25. Vertical slices of the (a) phase-contrast and (b) absorption-contrast nano-tomograms of the hydrated cement paste, UN: Unhydrated C3S, CH, C−S−H, Q (quartz capillary) and W (pore solution). (c) 3D renderings of the volume showing how the phases are located with respect to each other. (d) Bivariate histogram of absorption and phase. (e) Radial electron density profile of the particles as a function of distance from surface inward. The effective particle diameters are shown in the legend. (f) Site-specific water content of the C −S−H at the same slice shown in (a) and (b) with a resolution of 1 µm, where the colorbar is given in units of mol of water. Reprinted from reference 159 which is an open access article published under a Creative Commons Non-Commercial No Derivative Works (CC-BY-NC-ND) Attribution License. 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 Crystallography Reviews 39 large chemical variability. Advances in synchrotron tools will tackle these issues starting with the determination of chemical compositions and density values of these gels with very high spatial resolution by further development of the appropriate (combination) of imaging techniques. 2. The microstructure quantitative study of cement pastes is very important to understand and predict their mechanical behaviour as well as chemical durability. In this arena, synchrotron tools are very well suited as they do not require special sample preparation or sample environments that alter the microstructures. Here the challenge is to continue developing the imaging techniques, and the sample preparation procedures, for entry into the resolution range lower than 100 nm. A very good resolution, well below 100nm, without a trade-off of field-of-view is important to properly characterize key pore microstructure details: connectivity and tortuosity. It will be also important to quantify the changes in the microstructure provoked by the uses of SCM which can vary quite a lot (fly ashes, slags, partially burned clays, etc.). For this type of application, one of the techniques with the brightest future, in my opinion, is hard X-ray ptychographic forward coherent diffraction nano-tomography. 3. Most crystalline materials in cement chemistry have known crystal structures. However, some hydrates have still unknown crystal structures. For instance, C2AH8has been known for more than a century and its crystal structure is still not reported although it is known that it belongs to the AFm type structure. This is due to the combination of its chemical instability (it loses water very easily) with the lack of single crystals and that it is so far crystallized with additional coexisting phases. Here, microcrystal structure determination tools could be key to determine the crystal structure of this type of compounds from powder with grains smaller than 5 µm but using single-crystal-like techniques. 4. The footprint of cement production is high and should be reduced,but retaining the life standards. Therefore the development of eco-cements is important to decrease anthropogenic CO2emissions but maintaining the quality of our buildings and constructions. This can be tackled in a number of ways including the partial replacement of OPC by SCM but also by developing new binders, not based in OPC, like alkaline-activated materials and sulphobelite cements. In this direction, synchrotron techniques are being used, and will be used more intensively in the future, to shorten the time between chemistry formulation developments and their market appearances. This usage is very important as durability of new binder must be ensured and to do this, the understanding and quantification of their microstructures are vital. 5. Finally, it is worth mentioning that under-construction and planned diffraction-limited storage rings (fourth-generation synchrotron sources) will produce smaller beams with higher flux/brilliance of much higher coherence. These properties will directly impact the points described above and several others. The interested reader is directed to the special issue of Journal of Synchrotron Radiation published in September of 2014 which was fully devoted to the technical developments and science to be carried out in these last-generation synchrotrons. Acknowledgments I am grateful to all my coauthors, collaborators, colleagues and PhD students, for all our work together during more than two decades. I thank the University of Malaga and ALBA Synchrotron Light Source for the support and the stirring environments. I acknowledge the Spanish science funding agencies (they Q11 change the name quite often) for funding my studentship, to do the PhD and the three summer research stays at Oxford University, to the last ongoing research project. To all synchrotrons I have been allowed to Q12 enjoy carrying out experiments: SRS, ESRF, Max-Lab, DLS, APS, SLS and ALBA. Finally, this work has Q13 been supported by the Spanish MINECO through the BIA2014-57658-C2-1-R research grant. 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 40 M.A.G. Aranda Disclosure statement No potential conflict of interest was reported by the author. Notes on contributor Miguel A. G. Aranda received his Ph.D. from the University of Malaga (Spain) in 1992 in the field of Rietveld analysis for structure determination. During his Ph.D.he carried out three summer research stays at Chemical Crystallography Laboratory (University of Oxford). He carried out his postdoctoral training at the University of Cambridge with Paul Attfield to work on the crystal structure of Cubased high-Tc superconductors using synchrotron and neutron powder diffraction. He came back in 1994 as Assistant Professor to the University of Malaga where he also was Associate Professor and Professor. In January 2013, he moved to ALBA Synchrotron Light Source as Scientific Director. He has research experience in cements and building materials but also in other fields like: ceramics, pigments, cultural heritage and archaeometry as well as strongly electron-correlated transition-metal oxides, solid-oxide fuel cells and metal-organic-framework materials. His hobbies include walking in the hills, snorkelling, travelling and enjoying good wine and food. References [1] Mehta PK, Monteiro PJM. Concrete: microstructure, properties, and materials. New York: McGrawHill; 2013. Q14 [2] Damtoft JS, Lukasik J, Herfort D, Sorrentino D, Gartner EM. Sustainable development and climate change initiatives. Cem. Concr. Res. 2008;38:115–127. [3] U.S.: geological survey, mineral commodity summaries, January; 2009. [4] Taylor HFW. Cement chemistry. London: Academic Press; 1990. [5] Matschei T, Lothenbach B, Glasser FP. The AFm phase in Portland cement. Cem Concr Res. 2007;37:118–130. [6] Balonis M, Glasser FP. The density of cement phases. Cem Concr Res. 2009;39:733–739. [7] Richardson IG. Model structures for C-(A)-S-H(I). Acta Cryst. 2014;B70:903–923. [8] Willmott P. An introduction to synchrotron radiation. Techniques and applications. Chichester: John Wiley & Son; 2011. [9] Mobilio S, Boscherini F, Meneghini, C, editors. Synchrotron radiation basics, methods and applications. Berlin: Springer; 2015. [10] Schlachter AS, Robinson AH, Bienenstock A, Mills D, Shenoy G, Winick H. Synchrotron radiation. In: AccessScience. McGraw-Hill Education; 2014. Available from:http://www.accessscience.com/ content/synchrotron-radiation/675200 [11] Bertrand L, Robinet L, Thoury M, Janssens K, Cohen SX, Schöder S. Cultural heritage and archaeology materials studied by synchrotron spectroscopy and imaging. Cem Concr Comp. 2012;106:377–396. [12] Bertrand L, Cotte M, Stampanoni M, Thoury M, Marone F, Schöder S. Development and trends in synchrotron studies of ancient and historical materials. Phys Rep. 2012;519:51–96. [13] Helliwell JR. The evolution of synchrotron radiation and the growth of its importance in crystallography. Cryst Rev. 2012;18:33–93. [14] Patterson BD. Crystallography using an X-ray free-electron laser. Cryst Rev. 2014;20:242–294. [15] Monteiro PJM, Kirchheim AP, Chae S, et al. Characterizing the nano and micro structure of concrete to improve its durability. Cem Concr Comp. 2009;31:577–584. [16] Chae SR, Moon J, Yoon S, et al. Advanced nanoscale characterization of cement based materials using X-Ray synchrotron radiation: a review. Int J Concr Struct Mater. 2013;7:95–110. [17] Provis JL, Hajimohammadi A, White CE, et al. Nanostructural characterization of geopolymers by advanced beamline techniques. Cem Concr Comp. 2013;36:56–64. [18] Holt M, Harder R, Winarski R, Rose V. Nanoscale hard X-ray microscopy methods for materials studies. Annu Rev Mater Res. 2013;43:183–211. [19] De la Torre AG, Bruque S, Campo J, Aranda MAG. The superstructure of C3S from synchrotron and neutron powder diffraction and its role in quantitative phase analyses. Cem Concr Res. 2002;32:1347–1356. 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 Crystallography Reviews 41 [20] Petterson VK, Hunter BA, Ray A. Tricalcium Silicate T1and T2polymorphic investigations: rietveld refinement at various temperatures using synchrotron powder diffraction. J Am Ceram Soc. 2004;87:1625–1634. [21] Petterson VK. A Rietveld refinement investigation of a Mg-stabilized triclinic tricalcium silicate using synchrotron X-ray powder diffraction data. Powder Diff. 2004;19:356–358. [22] De la Torre AG, Lopez-Olmo MG, Alvarez-Rua C, Garcia-Granda S, Aranda MAG. Structure and microstructure of gypsum and its relevance to Rietveld quantitative phase analyses. Powder Diff. 2004;19:240–246. [23] Garver K, Beuchle G, Bornefeld M, Black L, Stemmermann P. Cell dimensions and composition of nanocrystalline calcium silicate hydrate solid solutions. part 1: synchrotron-based X-ray diffraction. J Am Ceram Soc. 2008;91:3005–3014. [24] Battocchio F, Monteiro PJM, Wenk HR. Rietveld refinement of the structures of 1.0 C-S-H and 1.5 C-S-H. Cem Concr Res. 2012;42:1534–1548. [25] Renaudin G, Filinchuk Y, Neubauer J, Goetz-Neunhoeffer F. A comparative structural study of wet and dried ettringite. Cem Concr Res. 2010;40:370–375. [26] Dilnesa BZ, Lothenbach B, Renaudin G, Wichser A, Kulik D. Synthesis and characterization of hydrogarnet Ca3(AlxFe1−x)2(SiO4)y(OH)4(3−y). Cem Concr Res. 2014;59:96–111. [27] Cuesta A, De La Torre AG, Losilla ER, Santacruz I, Aranda MAG. Pseudocubic crystal structure and phase transition in doped Ye’elimite. Cryst Growth Des. 2014;14:5158–5163. [28] Dilnesa BZ, Lothenbach B, Le Saout Get al. Iron in carbonate containing AFm phases. Cem Concr Res. 2011;41:311–323. [29] Runcevski T, Dinnebier RE, Magdysyuk OV, Pollmann H. Crystal structures of calcium hemicarboaluminate and carbonated calcium hemicarboaluminate from synchrotron powder diffraction data. Acta Cryst. 2012;B68:493–500. [30] Mesbah A, Francois M, Cau-dit-Coumes Cet al. Crystal structure of Kuzel’s salt 3CaO·Al2O3· $1/2CaSO4·1/2CaCl2·11H2O determined by synchrotron powder diffraction. Cem Concr Res. 2011;41:504–509. [31] Meral C, Benmore CJ, Monteiro PJM. The study of disorder and nanocrystallinity in C–S–H, supplementary cementitious materials and geopolymers using pair distribution function analysis. Cem Concr Res. 2011;41:696–710. [32] Skinner LB, Chae SR, Benmore CJ, Wenk HR, Monteiro PJM. Nanostructure of calcium silicate hydrates in cements. Phys Rev Lett. 2010;104:195502. [33] Soyer-Uzun S, Chae SR, Benmore CJ, Wenk HR, Monteiro PJM. Compositional evolution of calcium silicate hydrate (C–S–H) structures by total X-ray scattering. J Am Ceram Soc. 2012;95:793– 798. [34] Benmore CJ, Monteiro PJM. The structure of alkali silicate gel by total scattering methods. Cem Concr Res. 2010;40:892–897. [35] Mei Q, Benmore CJ, Sharma R, Yarger JL. Intermediate range order in vitreous silica from a partial structure factor analysis. Physical Review B. 2008;78:1–7. [36] Bell JL, Sarin P, Driemeyer PE, Haggerty RP, Chupas PJ, Kriven WM. X-ray pair distribution function analysis of a metakaolin-based, KAlSi2O6.5.5H2O inorganic polymer (geopolymer). J Mat Chem. 2008;18:5974–5981. [37] White CE, Page K, Henson NJ, Provis JL. In situ synchrotron X-ray pair distribution function analysis of the early stages of gel formation in metakaolin-based geopolymers. Appl Clay Sci. 2013;73:17–25. [38] Bell JL, Sarin P, Provis JL, et al. Atomic structure of a cesium aluminosilicate geopolymer: a pair distribution function study. Chem Mat. 2008;20:4768–4776. [39] White CE, Provis JL, Bloomer B, Henson NJ, Page K. In situ X-ray pair distribution function analysis of geopolymer gel nanostructure formation kinetics. Phys Chem Chem Phys. 2013;15:8573–8582. [40] White CE, Daemen LL, Hartl M, Page K. Intrinsic differences in atomic ordering of calcium (alumino)silicate hydrates in conventional and alkali-activated cements. Cem Concr Res. 2015;67:66–73. [41] Dilnesa BZ, Wieland E, Lothenbach B, Dahn R, Scrivener KL. Fe-containing phases in hydrated cements. Cem Concr Res. 2014;58:45–55. [42] Vespa M, Wieland E, Dahn R, Lothenbach B. Identification of the thermodynamically stable Fecontaining phase in aged cement pastes. J Am Ceram Soc. 2015;98. online. doi:10.1111/jace.13542 [43] Mendes A, Gates WP, Sanjayan JG, Collins F. NMR, XRD, IR and synchrotron NEXAFS spectroscopic studies of OPC and OPC/slag cement paste hydrates. Mater Struct. 2011;44:1773–1791. [44] Grangeon S, Claret F, Lerouge C, et al. On the nature of structural disorder in calcium silicate hydrates with a calcium/silicon ratio similar to tobermorite. Cem Concr Res. 2013;52:31–37. [45] Poo-arporn Y, Thachepan S, Palangsuntikul R. Investigation of damaged interior walls using synchrotron-based XPS and XANES. J Synchr Rad. 2015;22:86–90.