Improving the Sustainability of Cement and Concrete
Karen L. Scrivener
 |
In this paper I review options for improving the sustainability of cementitious materials. It is clear that blended cements based on Portland cement clinker with increasing levels of substitution are the most viable route. However the relative low amounts of commonly used SCMs mean that new sources are needed to continue this trend.Calcined clays are the only materials which are potentially available in large enough amounts to continue the trend of reducing clinker content in blended cements.
|
Fig.1 Traditional Site Mixed Concrete Under Usage in India - the most Primitive Methods of Concreting
Recently we have demonstrated that a combination of calcined clays with limestone can be used to achieve high levels of replacement of clinker in blended cements with good performance.
Fig. 2 Modern Eco Friendly RMC Plant with Advanced Manufacturing Process and technology fro Producing Quality Readymixed Concretes
The continued development of such blends and other materials with lower associated CO2 emissions requires a more scientific approach to understanding performance based mechanisms. In particular, so called “performance” tests for durability need to be more soundly based otherwise they will inhibit rather than facilitate the introduction of more sustainable materials. A co-ordinated research approach such as developed in the European network Nanocem, demonstrates that it is possible to progress rapidly our fundamental knowledge of the nanoscience of cementitious materials and so to meet the challenge of improving sustainability and meeting increasing demand.
Introduction
Cement and Concrete are essential to the infrastructure of the modern world. No other material is able to fulfil the growing demand for building materials with such a low environmental footprint. The widespread availability and low cost of cement make it by far and the most used material on earth, with reinforced concrete accounting to around half of all the “stuff” we produce. It is only because of these enormous volumes that overall the production of cement and concrete is estimated to account for around 5-8% of manmade CO2 emissions. For this reason, and with demand for cement forecast to more than double by 2050, there is growing pressure to improve the sustainability of this important material. In this case, all the pillars of the sustainability principle need to be considered: the impact on the environment in terms of CO2 emissions and raw materials; satisfying the needs of society for building materials; and doing this economically while maintaining the viability of the companies in the sector. The low cost and energy of production present a great challenge to improving sustainability in an economically viable way. This also means that cement and concrete must be local materials as the cost and environmental impact of transport can easily eliminate gains made elsewhere.
In this paper I will review options for improving the sustainability of cementitious materials and discuss the research needed to support the introduction of more sustainable materials.
Origin of Co2 Emissions
|
Table1: Embodied energy and associated CO2 emissions for common construction material %
|
|
Material
|
Embodied Energy ( MJ/Kg)
|
Carbon di oxide (Kg Co2 / kg)
|
|
Normal Concrete
|
0.95
|
0.13
|
|
Fired Clay Bricks
|
3.00
|
0.22
|
|
Road & Pavement
|
2.41
|
0.14
|
|
Glass
|
15.00
|
0.85
|
|
Wood(Plain Timber)
|
8.50
|
0.46
|
|
Wood(Multilayer board)
|
15.00
|
0.81
|
|
Steel(from ore)
|
35.30
|
2.83
|
|
Steel(recycled)
|
9.50
|
0.43
|
Table 1 shows relative figures for the energy and CO2 emissions of some common building materials(1). Typical figures for concrete are shown here as this is the final product used in the field. Nonetheless, as discussed below, there is a huge scope also for reducing the amount of cement in concrete and it should always be borne in mind that sustainability can and should be considered at all stages of the process. Of course the figures in Table 1 are per kg and the strength of concrete is not the same as the specific strength of steel or wood. Such issues have been addressed by several authors e.g. [2] and the advantages of concrete are still clear in most construction related applications. Beyond this, the availability of materials to substitute concrete also needs to be considered. For example in 2005 it is estimated that the world production of reinforced concrete was around 17 billion tonnes (GT) while that of wood was around 2 GT – about one tenth. It is already considered that our consumption of wood is unsustainable; we are cutting down more trees than we are planting. So independently of any other consideration, wood cannot substitute for concrete to any significant degree.
Cement production accounts for the overwhelming majority of the CO2 emissions associated with concrete. Unlike other materials, only a minority of emissions from cement production are related to fuel and electricity (~40%). The remaining 60% comes from the decomposition of the main raw material –limestone, or CaCO3. On the energy side great improvements have been made in the last few decades. The production of Portland cement clinkers is one of the most efficient industrial thermal processes in existence (up to 80% of theoretical efficiency) and it is unlikely that significant further gains can be made here and discussed in more detail by Gartner [3]. Furthermore cement plants can now use a wide range of substitute fuels so that the need for primary fossil fuels may be below 20% in a modern plant. The versatility to exploit the calorific value of such a wide range of wastes in a safe manner should be seen as another advantage of cement production.
Therefore we need to focus our attention on the 60% of “chemical” emissions coming from the decarbonation of limestone. However, lowering emissions here has the inevitable consequence of changing the cement chemistry. This in turn means that the reactions and performance of new materials will not necessarily be the same as the reference Portland cement. This means that the whole basis for usage has to be re-established. The basis for use of reference Portland cement has taken more than 100 years of empirical testing to develop. The use of the most common blended materials – blends of Portland cement clinker with slag and fly ash has taken more than 30 years to become established. We do not have the time to go through this long testing phase for every new material which comes along, so we must move towards a more scientific basis for understanding and performance tests for users to have confidence in the many potential solutions. This can only come (on a reasonable timescale) through a systematic, understanding of cementitious processes and materials at the nanoscale; extended across all the scales involved in cement and concrete production to provide the multidisciplinary assessment and prediction tools needed to assess the functional and environmental performance of new materials.
Possible Chemistries
If we intend to reduce the use of calcium carbonate, then it must be replaced by something else. Again Gartner has discussed the different options in detail [3]. Simplistically we can consider the overall composition of the earth’s crust. Just 8 elements make up more than 98% of this – oxygen, silicon, aluminium, iron, calcium, sodium, potassium and magnesium. The relative abundances are shown logarithmically in Figure 1 [4] and, as a first approximation, cost will be closely related to availability. So really we need to consider only 7 oxides as possible candidates for cement making. These are summarized in Table 2, which roughly considers first their potential to form space filling hydrates and then their geological distribution.
The potential for space filling hydrates is the key to hydraulic cement. The anhydrous cement is mixed with water and the solid content cannot exceed around 50% (W/C ~ 0.3) if we are to have a flowable mixture. Then the anhydrous material must dissolve and precipitate new solids incorporating water with a higher solid volume than the original anhydrous material. The phases present in Portland cement clinker are ideal in this respect as there is roughly a doubling in solid volume, which enables most of the originally water filled space to be filled to give a strong solid with low porosity. If we consider the alkali oxides – Na2O and K2O – these alone cannot produce hydraulic compounds as these have a very high solubility and will not deposit hydrates. This means that the low amounts present in Portland cement end up nearly all in the pore solution, giving the high pH needed to protect reinforcement. If we now consider iron and magnesium oxide, we have the opposite problem. These do form insoluble hydrates, but their mobility in the alkaline solution of cements is very low, which means these hydrates are nearly exclusively deposited within the boundaries of the original grains and do not contribute to filling the originally water filled space or binding the grains together. Even in very old cement pastes (even > 100 years [5]), we see bright areas corresponding to the original ferrite phase. Slags, which contain significant amounts of magnesium, hydrate to give dark inner product rims, where all the hydrocalcite phase containing the magnesium is located. So this brings us back to the three oxides which dominate Portland cement – lime, silica and alumina.
Based on these most important oxides we can envisage three main routes to the development of more sustainable materials, which will be discussed in more detail below:
- Alternative hydraulic minerals from which to make alternative clinkers
- Portland cement clinker blended with supplementary cementitious materials (SCMs) containing these three main oxides
- Clinker free materials, notably SCMs activated by alkalis
Alternative Clinkers
Figure 2 shows the “chemical” CO2 emission of the hydraulic minerals in the CaO – SiO2 – Al2O3 – (SO3) system, from two view points, first per gram of anhydrous material, then per ml of hydrates which can be formed as it is this later which is most related to their space filling capacity and so to strength development. The volume of hydrates for the aluminate compounds (monocalcium aluminate and ye’elimite) is shown with and without added sulfate. Here it can be seen that the figures for the two most important Portland cement clinker mineral – C3S and C2S have fairly similar values. Notably we see that C2S only presents a saving of around 10% in CO2 compared to C3S. This means that for belite rich cements to represent a “real” CO2 saving, they have to have better then 90% of the performance of alite rich cements. Despite intensive research efforts, we are far from achieving this in terms of the performance needed in modern construction. This table clearly shows that the most interesting minerals from the point of view of their associated CO2 values are monocalcium aluminate (CA), the main component of calcium aluminate cements(CACs) and ye’elemite or Klein’s compound (C4A3$) the main component of calcium sulfoaluminate cements (CSAs of SACs). So it is worth discussing the viability of clinkers based on these minerals in some more detail.
The main problem is that while alumina is very abundant in the earth’s crust it exists mainly in feldspars and clays, where it is present alongside silica with Si:Al~ 2 similar to that found in most Portland clinkers. To obtain the desired alumina rich compounds a more concentrated source of alumina is needed, typically bauxite. But bauxite is nothing like as widely distributed as clay minerals and is highly sought after for the production of aluminium so this is a much more costly raw material. But this is not the end of the story, as will be explained from the perspective of calcium aluminate cements.
Calcium aluminate cements (CACs) have been in commercial production for nearly 100 years, the total amount produced annually is much less than 1/1000 of the amount of Portland cement clinker production and only a small proportion of this is used in construction. This is because the cost of CAC is more than five times more than the cost of Portland cement. The higher cost of raw materials and production account for only part of this huge discrepancy. The other part lies in the number of people needed to support the use of this special material, in terms of factories producing smaller amounts, sales and technical assistance for specialised materials which does not behave in the same way as the reference Portland cement. Consequently the use of CAC can only be justified where their special properties justify the higher cost.
Calcium sulfo aluminate cements have gained a lot of attention recently, as, in addition to the low chemical CO2 of ye’elemite the firing temperature is low and they are more easily ground. CSAs are a very diverse family of cements. They may contain from 70% ye’elemite down to 30% or less. The main other phase present is usually belite and this leads to a two stage development of properties, with the fast reaction of the ye’elemite giving early strength and the belite hydration giving later strength gain. Another complication is that the behaviour can be changed from rapid hardening to expansive by the amount of calcium sulfate added during grinding. Due to this diversity it is difficult to generalise, but it can certainly be said that today they are specialist products along the lines of CACs rather than cements for general construction. They have been produced in China for several decades and were used for some large scale constructions, but they do not today represent a significant part of the general construction market in China compared to Portland cement. There are a number of other companies producing different CSAs, mostly sold as rapid setting or expansive cements.
However, I am doubtful that CSA cements can be competitive with blended cements discussed below on either an economic or environmental basis in general use. The need to support consumers in the use of cements which behave differently to the Portland reference, in addition to the higher raw materials cost will mean that the commercially viable cost will be significantly more than that of Portland cement. Nevertheless such developments may be interesting in more specialist applications, for example precast concrete.
Blended Cements
The discussion above about alternative clinkers really emphasises the fact that Portland cement clinker is a highly optimised product. It can be produced in highly efficient factories producing several million tonnes per year from widely available and cheap raw materials. Production on a large scale on a continuous basis also has the great advantage of giving a constituent product with predictable performance. By and large, Portland cement is a very forgiving, robust product which can be used successfully by people with little technical knowledge.
For this reason the most successful strategy to improve the sustainability of cement up to now has been to blend Portland cement clinker with other materials, known generically as supplementary cementitious materials (SCMs). These materials are often waste or by-products from other industries, so there is a double benefit of usefully recycling these materials rather than sending them to land fill in addition to reducing clinker consumption. Although there are reports of SCM additions from the early part of the 20th century, this technology only started to be taken up on a large scale from the 1980s and received a major impulse with the introduction of the European Standard, ENV 197 allowing blends with 5-95% substitution of clinker according to five main types. Figure 4 shows how the clinker content of the cements produced by Holcim group
companies has evolved since 1990. It can be seen that over this time scale the clinker factor has been reduced by over 10% (roughly equivalent to 10% saving in CO2 emissions). Over the same time period the relative amount of different cement types has also evolved dramatically as shown in Figure 5. Most notably the proportion of ordinary Portland cement (OPC) type (>95% clinker) has decreased from over half to under a quarter of the cements sold. However, it can also be noted that there has been relatively little evolution over the past 3-4 years. This indicates that the present technology is reaching a limit due to: first the availability of common SCMs, and second the performance of these blended materials. I will return to these points later, but first it is useful to look at the most commonly used SCMs (slag, fly ash and limestone) in a bit more detail.
Slag
“Slag” as used in cements is a by-product of the production of pig iron in a blast furnace. Small droplets are rapidly quenched to produce a glassy material with a high amorphous content. Slag is a hydraulic material, that is to say it will react on its own with water, without the need for calcium hydroxide, albeit slowly. The reaction is accelerated by the alkaline pore solution of cement paste and blends of slag with Portland clinker are used with up to 95% slag and commonly in the range 40-70%. Slag is an excellent
SCM as it can be produced with consistent quality, rapidly catches up with OPC in terms of strength development and blends show excellent durability in environments containing chloride or sulfates. The major drawback is that the worldwide production of slag is only around 5% of the production of clinker. The vast majority of blastfurnace slag produced is already used in cement and concrete and as a material for use in concrete the price can be almost as high as that of Portland cement. So despite the many advantages of this material it cannot help much to further improve the sustainability of cementitious materials.
Flyash
Flyash comes from the incombustible residue to coal burnt to produce electricity. While the sustainability of electricity production from coal can itself be questioned, this will remain a major source of energy for at least the next 50 years, especially in countries like China and India. Worldwide the amounts of flyash produced are much higher than the amount of blast furnace slag, but still perhaps only one third of the amount of clinker produced. Furthermore the quality of flyash is highly variable, which means that much are not suitable for cement production. There are efforts to improve the quality of flyash with respect to its use in cement, by classification of the finer particles and even by additions of ingredients to change the chemistry, but the quality of the fly ash will always tend to be more dependent on the economics and regulations of power production and is suitability as an SCM.
Flyash is a pozzolanic material, which means that it must reacts with calcium hydroxide from the clinker phases to produce C-S-H. This requirement for calcium hydroxide means there is a limit to the amount of flyash which can react in a blend. Of course, this limit varies with the composition of the flyash and its reactivity, but the example presented in [7] shows that for a typical class F flyash, reacting 50% it is in the range of 20% addition. Levels of addition beyond this will just be inert filler. Of course, as discussed below for limestone, higher levels of addition may still be OK for general use if the strength requirement is adequate. Flyash also reacts much more slowly than the clinker phases with strengths of fly ash blends only catching up with those of the reference OPC after more than 28 day, and often not until more like 90days (Figure 6). This also means that in blends it is much more vulnerable to lack of proper curing.
So, while flyash makes and will continue to make an important contribution to more sustainable cementitious materials, there are unlikely to be large further breakthroughs from this route.
Limestone
The biggest selling cement types in Europe contain limestone as an SCM, CEMII-L and CEM II-M types. Limestone is readily available at the cement plant, as it does not pass through the kiln the “chemical” CO2 is zero and, it is generally easily ground, consuming very little energy. From these perspectives it is an excellent SCM.
For many years there was controversy as to whether limestone reacted in cement pastes. Thermodynamic studies [8, 9, 10] now show clearly that small amounts can react with available aluminate to give calcium monocarbo and hemicarbo aluminate. This reaction means that the overall volume of hydrates can be increased – with a consequent gain in strength for additions in the range 3-5% to a typical Portland cement. In addition when limestone is interground with clinker the fine particles then act as extra nucleation sites for C-S-H during early hydration accelerating the hydration of the clinker component. For these reasons cements with up to 5% limestone show better performance in most respects compared to pure Portland and this level of addition is common in most CEM I types in Europe and many other countries. At higher levels of addition the limestone acts as an inert filler. The fine particles generally improve rheology so cements with 10-20% limestone can be very cost and environmentally effective for concrete where only normal strength levels are required and the service environment is not aggressive. Another advantage of the addition of limestone at the grinding stage is that lime stones containing dolomite, not suitable for clinker manufacture, can be used [11, 12], extending the life of quarries.
Limitations On Reducing Clinker In Blended Cements
From the above discussion it is clear that blended cements are being used more and more worldwide. We now have extensive experience of these materials to see that they can perform well over the life of buildings and structures. Another important factor in the uptake of these materials is that their handling is very similar to Portland cement so they can generally easily be used with a minimum of adjustment in mixing, placing, etc. However, increasing the level of clinker substitutions is limited by two important factors:
1. Adequate strength at early ages: The characteristic strength of a concrete is usually taken at 28 days, but in practice the choice of concrete to use is often determined by the strength it achieves at much younger ages, even as little as 2-3 days. Most SCMs make little or no contribution to strength at such early ages, as illustrated by Lawrence et al [13].
2. Supply of SCMs: as indicated above the amounts of common SCMs available are small in relation to the quantities of clinker produced. Many alternative SCMs have been investigated, but none of these even approaches the quantities of slag, with the exception of calcined clays and perhaps natural pozzolans.
Calcined Clay / Limestone Blends
The pozzolanic activity of calcined clays has long been known. Kaolinite is by far the most reactive clay when calcined and relative pure “metakaolins” are available commercially, but at a price 2-3 times that of cement. It has been shown that relatively impure clays, with as little as 40% kaolinite content, can also give excellent strength in cement blends [14]. Such low grade clays are widely available throughout the world. Nevertheless, the need for calcination presently puts these materials at a disadvantage compared to SCMs which are waste products of other industries, even though calcination does not produce “chemical” CO2 and the optimum calcination temperatures of 600-800°Care very much lower than the temperatures needed to produce clinker.
Recently we demonstrated that co-substitution of clinker with a combination of calcined clay together with limestone, can give good mechanical performance as early as 7 days at substitution levels of 45-60% [15]. In Figure 6 the strength at 7 days of these blends is compared to common SCMs and inert quartz as studied by Lawrence et al.
Alkali Activated Materials
At present many researchers are investigating alkali activated materials, also known as “geopolymers” claiming the advantages of these as low CO2 cements. Despite the hype, it is highly unlikely that such materials will have a major impact on the sustainability of cementitious materials for many reasons:
• First the environmental advantages of these materials are questionable. The environmental impact made by Habert et al [16] indicates that while the production of geopolymer concrete can have a slightly lower impact on global warming than standard ordinary Portland cement (OPC) concrete, it has a higher environmental impact regarding other impact categories related to the production of concentrated sodium silicate solution and the dangers of handling highly alkaline solutions.
• Second, the performance of these materials is very sensitive to small variations in the starting materials, such as slag or flyash. The setting time and rheology are hard to control and may vary widely on a batch to batch basis.
• Third, and perhaps most important, the materials used in these alkali activated materials are the same SCMs discussed above – slag, flyash, etc., which as we have seen are only available in relatively low amounts compared to cement clinker. The ones which perform best in these systems are the same as those which perform best in cement blends.
Research Needs: Nanocem Roadmap
From the above discussion, it should be clear that there is no single “magic bullet” solution to sustainability. Potential solutions must be adapted to the local resources and conditions and whatever they are, they can only be successful if we can provide the basis in understanding and performance tests for users to have confidence in the many potential solutions. This can only come (on a reasonable timescale) through a systematic, science-based under-standing of cementitious processes and materials at the nanoscale; extended across all the scales involved in cement and concrete production to provide the multidisciplinary assessment and prediction tools needed to assess the functional and environmental performance of current and new materials.
Given below are just a couple of examples of ongoing work and the questions to be answered.
Hydration Kinetics: Limitations to Strength Development
The general hydration kinetics of Portland-based cements are well known. On the addition of water, the hydration of cement is initially very rapid, but slows quickly. This period of slow reaction, the induction period, is essential in the practical use of cement as it remains fluid and can be transported and placed. This period is thought to be controlled by the dissolution kinetics and under saturation of the aqueous environment. At the end of the induction period the reaction accelerates with a maximum rate around 10
hours, and then slows and is very low at 24 hours. The setting of cement occurs during the acceleration period and strength develops rapidly, but slows down at the onset of the deceleration period. Delaying the onset of the deceleration period would provide higher early strength and allow higher levels of clinker substitution (lowering environmental impact). During the acceleration period the hydration kinetics are thought to be controlled by the nucleation and growth of the poorly ordered C-S-H phase. The reasons for the onset of the deceleration period are not clear and need further investigation.
Determining Resistance to Carbonation
There is no consensus in the literature on the reliability of accelerated tests. Recently we investigated the carbonation of cements with low clinker content in both accelerated and normal conditions. It was found that accelerated carbonation in an environment with only 4% CO2 does not describe properly the actual
carbonation pattern of the blends. It can be explained by the formation of different calcium carbonate polymorphs in the blends for accelerated carbonation that are not representative of the reality in real carbonation conditions [18]. With the absence of portlandite and their lower carbonatable content, such accelerated tests may be unjustifiably negative with respect to blends with low clinker content. Carbonation in concrete specimens made with blends exposed in natural conditions shows good performance in comparison with reference specimens made with Portland cement, both cured in similar conditions. Thomas has reported long term data from fly ash concrete, which shows that the carbonation depths of similar strength materials are almost identical after 10-20 years [19].
Concluding Remarks
In this paper I have tried to outline the realistic options for lowering the environmental impact of cementitious materials. Routes, which can really have a significant impact in practice must be based on widely available raw materials and robust technologies not significantly different from those used with Portland cement concrete today. The past few decades have proved that cements based on Portland cement clinker substituted by supplementary cementitious materials (SCMs) are the only really viable options to lower environmental impact. To go further we need to find new sources of SCMs and here calcined clays look to be by far the most promising materials. Calcined clays, combined with limestone allow much higher levels of substitution while maintaining good strength development from a few days.
They are widely available worldwide and the most reactive clays, containing high proportions of kaolinite are found in countries where demand for cement is expected to increase most strongly in coming years. Consequently such materials are promising not only to reduce environmental impact but also to satisfy the growing demand for building materials, which can only be realistically met by concrete.
To successfully bring materials with a lower environmental impact to the market, we need to better understand the fundamentals controlling first the hydration reactions and then durability. There is currently much discussion about performance testing. In principal this is a good idea to allow faster implementation of alternative materials. However, most accelerated test methods being promoted have been developed for pure Portland cements or blend with low levels of substitution. There is increasing evidence that these tend to be unrealistically negative for materials with high levels of clinker substitution. Adopting methods without a proper understanding of the underlying mechanisms, risk the introduction of the impeding new materials – exactly the opposite of their intention.
Acknowledgements
I am very grateful to many colleagues, mainly within the Nanocem consortium, for discussions leading to the subject matter of this paper. However, opinions expressed are my own and not necessarily endorsed by all members of the consortium.
Author
Karen L. Scrivener, Laboratory of Construction Materials, Institute of Materials Ecole Polytechnique Fédérale de Lausanne, EPFL, Lausanne, Switzerland
References
1. Hammond G.P. and Jones C.I, ICE version 1.6ª, 2008 Proc Instn Civil Engineers. www.bath.ac.uk/mecheng/sert/embodied/
2. P. Purnell, Material Nature versus Structural Nurture: The Embodied Carbon of Fundamental Structural Elements, Environmental Science and Technology, 2012, 46, 454–461, dx.doi.org/10.1021/es202190r |
3. E.M. Gartner, Industrially interesting approaches to ‘‘low-CO2’’ cements, Cement and Concrete Research, (2004) 1489 – 1498
4. http://upload.wikimedia.org/wikipedia/commons/0/09/Elemental_abundances.svg
5. Scrivener, K.L., "A study of the microstructure of two old cement pastes", 389-393. (Proc. 8th Int Cong. on the Chemistry of Cements, Vol. III, Rio de Janeiro, 1986)
6. http://www.holcim.com/sustainable-development/performance-data/tables-and-graphs.html
7. Barbara Lothenbach, Karen Scrivener, R.D. Hooton, Supplementary cementitious materials:
Review, Cement and Concrete Research, 41(12) 1244-1256, December 2011
8. Lothenbach, B., et al., Influence of limestone on the hydration of Portland cements, Cement and Concrete Research, 2008. 38(6): p. 848-860.
9. Matschei, T., B. Lothenbach, and F.P. Glasser, The role of calcium carbonate in cement hydration.
Cement and Concrete Research, 2007. 37(4): p. 551-558.
10. Matschei, T. and F.P. Glasser, The influence of limestone on cement hydration. Zum Einfluss von Kalksteinauf die Zementhydratation, 2006. 59(12): p. 78-86.
11. Zajac M., Dienemann W. Bolte G. Comparative experimental and virtual investigations of the influence of calcium and magnesium carbonates on reacting cement, Proceedings 13th ICCC,
Madrid 2011
12. Schöne S., Dienemann W, Wagner E, Portland dolomite cement as alternative to Portland limestone cement, Proceedings 13th ICCC, Madrid 2011
13. P. Lawrence, M.Cyr, E. Ringot, Mineral admixtures in mortars effect of type, amount and fineness of fine Constituents on compressive strength, Cement and Concrete Research 35 (2005) 1092–1105
14. Fernandez Lopez, R., Calcined clayey soils as a potential replacement for cement in developing countries, Thesis EPFL, 2009
15. M. Antoni, J. Rossen, F. Martirena, K. Scrivener, Cement substitution by a combination of metakaolin and limestone, Cement and Concrete Research 42 (2012) 1579–1589
16. Habert, G., d’Espinose de Lacaillerie, J.B., Roussel, N. 2011. An environmental evaluation of geopolymer based concrete production: reviewing current research trends. Journal of Cleaner Production, 19, 1229-1238
17. Nanocem: The industrial-academic nanoscience research network for sustainable cement and concrete, www.nanocem.org
18. Antoni, M., Investigation of cement substistution by blends of calcined clay and limestone, PhD Thesis, 2013, EPFL: Lausanne
19. M.D.A. Thomas, Supplementary Cementitious Materials in Concrete, CRC Press, 2013