The Use of MgO by the Chinese
The Chinese have been beneficially using magnesia in concretes, particularly for Dam manufacture since the 1970’s and we have often referred to the fact in support of our own arguments. They add a moderately reactive magnesia so that "long-term autogenous expansion at later ages compensates
for contraction due to cooling and shrinkage.". TecEco on the other hand use a much more reactive grade of magnesia with the objective of getting as much Mg++ into solution as quickly as possible so as to also improve early age properties such as rheology and plastic shrinkage.
We mention the Chinese contribution when discussing the history of magnesium in hydraulic cements at http://www.tececo.com/history.magnesium_hydraulic_cements.php and in relation to calcination temperature in our discussion about Reactive Magnesia. Note also that the history page also mentions magnesia in Roman and Rosendale cements as well. For those further interested an excellent paper summarising the Chinese use of magnesia in concretes by Du appeared in the December issue of Concrete International.
The form of the magnesia the Chinese have been using is quite different to the form TecEco have been using and the hydration rate and effects are therefore also quite different. Our specification makes it clear that we prefer as low temperatures as technically possible for calcination and our maximum temperature is 50 degrees C lower than the minumum used by the Chinese according to Du. On page 9 of our original PCT application (AU01 00077) at lines 4-8 “Suitable magnesia should be calcined at low temperatures (less than 750°C) and ground to greater than 95% passing 120 micron. Generally the lower the temperature of calcination and finer the grind, the more reactive the magnesia is and the faster it hydrates. Magnesia calcined as 650 °C passing 45 micron or less is better.”
The main reason we specify a much more reactive magnesia than the Chinese have in the past is that we wanted magnesia to rapidly go into solution for the effect the highly kosmotropic magnesium ion has on early age properties of fresh concrete such as rheology, bleeding, gelling and other characteristics as detailed on our web site at http://www.tececo.com/technical.rheological_shrinkage.php.
In contrast the main reason the Chinese have been using magnesia is that the delayed hydration compensates for autogenous and well as drying shrinkage.
Although there may be some autogenous expansion of magnesia depending on where the water required is coming from TecEco mainly resolve both drying and autogenous shrinkage by controlling loss of volume through reduced bleeding and evaporation and autogenous shrinkage by allowing without distress the more complete hydration of PC and magnesia from polar bound water within the system in brucite hydrates. (See http://www.tececo.com/technical.rheological_shrinkage.php.).
In relation to our patents on the bottom of page 8 of our original PCT application we explained the importance of hydration rate “The key for the successful blending of magnesia and other cements and in particular Portland type cements is that the hydration rates of all components in the cement must be matched. In order to achieve this the magnesia component must be separately calcined at lower temperatures and in conditions that are suitable for the manufacture of reactive magnesia”
Magnesia that hydrates to brucite in the same time frame as other components in a hydraulic cement mix can only be made by low temperature calcination whereas in the concretes used by the Chinese the hydration of magnesia must be significantly delayed to compensate for shrinkage by expansion and this is achieved by manufacture at higher temperatures.
Numerous authors including Birchal et. Al. in his conclusions on page 1632 and Blaha et. Al in parts 2  and 3  also make it clear that the temperature of firing is all important explaining that the lower the temperature of calcination, the more reactive the magnesium oxide will be and the faster the rate of hydration.
Du referred to earlier  shows on page 47 a table reproduced below that serves to indicate the enormous effect the calcination temperature of MgO has on the hydration rate expressed as the time taken for full hydration which is an important indicator and outcome of greater reactivity.
Hydration of MgO Powder from Du, C. (2005).
The maximum temperature of calcination according to our specification is 750 deg C which is at least 50 deg C lower than the minimum in the table and used by the Chinese and 100 deg C less than the lowest temperature considered as lightly burned according to the meaning of the term given by Du and the authors he has cited.
The importance of temperature of calcination for reactivity is also made very clear in “Gelling Materials Science”, 1st Edition, pages 43-50  of which we have a translation we can publish if necessary and particularly in an article by Dell and Weller that appeared in the Transactions of the Faraday Society in 1959.
Fig 2 from Dell and Weller confirming the Optimal Maximum Temperature of Calcination for Reactivity is less than 750 degrees C 
Figure 2 from Dell and Weller above shows quite clearly that the optimal maximum temperature for highly reactive material is about 750 degrees C, at least 50 degrees cooler than the minimum used by the Chinese. Our preferred temperature 650 degrees C is 150 degrees C cooler .
Du refers to the degree of crystallinity being the relevant factor. We prefer to describe reactivity mainly in terms of a kinetic barrier being the lattice energy of the magnesia. Lattice energy is not easily measured; in fact it cannot be directly which is probably why many papers refer to specific or internal surface area. Specific surface area is a measure of the reactivity of magnesium oxide and is related to the lattice energy and grind size, defects etc. The higher the specific surface area the lower the lattice energy. The lower the lattice energy barrier to be overcome, the greater the hydration rate.
That 750 degrees C is an optimal upper temperature for highly reactive magnesia that will rapidly go into solution affecting early properties of a concrete mix is further supported by Fig 7 from Dell and Weller  below which demonstrates that there is no further loss of either moisture or CO2 above about 750 degrees C.
In his paper on page 47 Du makes it clear that the Chinese used MgO decomposed in the range 800 – 850 degrees C which results in hydration rates that are slower and that compensate for drying and autogenous shrinkage by expansion. On page 47 Du states “When hydrated, the volume of the Mg(OH)2 crystals formed from MgO is significantly larger than the sum of the volumes of MgO and H2O alone, resulting in autogenous expansion”.
TecEco use more reactive magnesia so that hydrates in the same time frame as other components of a hydraulic mix and therefore obtains water from excess water in the mix and does not result in the same expansion through later stoichiometric expansion as is the mechanism with the Chinese use of magnesia. In this way other early age properties are affected such as plastic shrinkage, rheology and bleeding as set out on the TecEco web site on the web page titled Rheological and Shrinkage Reduction Affects of Adding Reactive Magnesia to Concretes.
In TecEco cements expansion is also controlled by maintenance of whole system volume - less bleeding and less evaporation and higher surface tension, the enthalpy of evaporation and surface tension being increased by the presence of the strongly kosmotropic magnesium ion. The extent to which shrinkage is controlled by delayed expansion depends entirely on where the water is coming from.
When magnesia hydrates it may or may not be expansive depending on whether the water used comes from within or from outside the system cement + aggregates + water:
MgO (s) + H2O (l) = Mg(OH)2 (s)
40.31 + 18.0 => 58.3 molar mass
11.2 + liquid => 24.3 molar volumes
Solidus expansion can be up to 116.96% depending on whether the water is coming from mix water, bleed water that would otherwise have exited the system or from outside the system.
By adding highly reactive magnesia most of the water for hydration comes from mix water or excess water that would otherwise have bled out of the system. The lack of bleed water in TecEco concretes means there is less early age shrinkage and more polar bound water for later more complete hydration of PC and magnesia without dimensional distress due to stoichiometric or autogenous shrinkage - the main cause of dimensional distress after loss of water that tends to occur later.
It follows by varying the amount and form of magnesia very good dimensional control can be achieved. In the case of TecEco by controlling the whole system volume by preventing losses and by preventing stiochiometric shrinkage through the supply of polar bound water and in the case of the Chinese by matching inevitable shrinkage with balanced expansion. The varying factor is the rate of hydration which we explain above depends on lattice energy and the temperature of firing.
It is important to notice how steep the curves in both of the above references are. A 750 degrees C max with 650 degrees C preferred calcination temperature results in vastly different reactivity to that produced at 800 – 850 degrees C used by the Chinese
The importance of calcination in relation to hydration rate is also confirmed independently by Blaha. who examined in detail the affect of conditions of calcination on hydration rate. Blaha states in the abstract to the above paper “The specific surface area of the oxide decreases exponentially with increasing firing temperature.” He produces the graph below at page 22. Note the previously discussed relationship between specific surface area and lattice energy.
Figure 1 from Blaha. Specific surface area of MgO vs temperature of firing. Time of firing is 60 minutes.
TecEco freely acknowledge the pioneering work of the Chinese. Our contribution was essentially that temperature was very important and that if reactive enough magnesia will hydrate in the same time frame as Portland cement and many other hydraulic cements and could therefore be used to also affect the properties of fresh concrete and be blended with it in any proportion. The Chinese on the other hand were using the delayed expansion of magnesia to control shrinkage and did not recommend over 5% addition.
Du above makes it clear in his text that fine grinding is important in contrast to our teaching which is that proper particle packing depending on purpose is what matters (See The Importance of Particle Packing). The reason the Chinese emphasise fine grinding is because the more finely and well dispersed their magnesia is then the less likely there is to be localised dimensional distress as its expansion counteracts shrinkage.
Our large particle size maxima (greater than 95% passing 120pm) in our specification but relatively low temperature of calcination were deliberately chosen because the water demand for larger particles is less in a hydraulic cement and it is well known in the industry and expressed by Duff Abrams "law" that strength is inversely proportion to the amount of mix water. What matters however in relation to water demand is not so much the size of particles but how they pack. See The Importance of Particle Packing and we use both larger and smaller particles in order to pack them properly.
By using much more reactive magnesia that the Chinese, not only are autogenous expansion, the modulus of elasticity, tensile strain capacity and creep favourably affected, but many early age properties such as rheology, plastic shrinkage and bleeding are also improved.
 Du, C. (2005). "A Review of Magnesium Oxide in Concrete - A serendipitous discovery leads to new concrete for dam construction." Concrete International (December 2005): 45 - 50.
 Birchal, V. S. S., S. D. F. Rocha, et al. (2000). "Technical Note. The Effect of Magnesite Calcination Conditions on Magnesia Hydration." Minerals Engineering 13(14-15): 1629-1633
 Blaha, J. (1997). "Kinetics of Hydration of Magnesium Oxide in Aqueous Suspension, Part 2. The Effect of Conditions of Firing Basic Magnesium Carbonate on the Specific Surface area of Magnesium Oxide." Ceramics - Silikaty 41): 21-27.
 Blaha, J. (1997). "Hydration Kinetics of Magnesium Oxide, Part 3. Hydration Rate of MgO in terms of Temperature and Time of Its Firing." Ceramics - Silikaty 41(4): 121 - 123.
 CBMA Gelling Materials Science.
 Dell, R. M. and S. W. Weller (1959). "The Thermal Decomposition of Nesquehonite MgCO3 3H20 And Magnesium Ammonium Carbonate MgCO3 (NH4)2CO3 4H2O." Trans Faraday Soc 55(10): 2203 - 2220.