Low Temperature Synthesis of High Alumina Cements by Novel
Low Temperature Synthesis of High Alumina Cements by Novel
, 1–8 (2013)
DOI: 10.1111/jace.12403
© 2013 The American Ceramic Society
Journal
Low Temperature Synthesis of High Alumina Cements by Novel
Co-Melt Precursors and Their Implementation as Castables with
Some Micro Fine Additives
Vijay Kumar,† Vinay Kumar Singh, and Abhinav Srivastava
Department of Ceramic Engineering, Indian Institute of Technology (BHU), Varanasi 221005, Uttar Pradesh, India
In the present investigations nano size high alumina cements 70–80 wt% Al2O3 can be used to produce castables with a
(HAC) were prepared by very effective co-melt precursor temperature resistance ranging from 1800°C to 1900°C.3
sintering technique from their metal nitrate precursors. The These materials are used for many special applications in the
prime cementing phases observed were CA, CA2, and C12A7. construction, civil engineering, and refractory industries
The addition of nano structured cements in refractory castables because of their ability to gain strength rapidly and to with-
has improved the thermo-chemical-mechanical properties to a stand aggressive environments and high temperature.4,5 The
significant extent. Each batch of low cement castables (LCC) knowledge of their mechanical behavior at a young age
was prepared from calcined Chinese bauxite, HAC, and super- becomes essential to forecast their performance in service.
fine additives. The effect of HAC in bauxite castable with the To describe hydration of cement paste at the early age (from
additives similar to Silicon Carbide, reactive alumina, and a few minutes to a few hours after mixing), different
micro-fine silica on the sinterability and properties of these approaches have been developed. Conductivity and pH
castables was investigated. Physical properties such as appar- measurements, usually carried out on dilute systems have
ent porosity and bulk density, mechanical properties such as been applied to predict the dissolution and precipitation
hot modulus of rupture (HMOR), cold and hot modulus of processes.6,7 Nano ceramic materials (<100 nm grain size) are
rupture (CMOR), and cold crushing strength (CCS) of still in the early stages of development, but depict many pro-
hydrated and sintered castables were studied. The sintered cessing and property advantages over their conventional
castables were also characterized for their solid phase composi- coarse grained alternatives. In recent years there has been
tions and microstructure using X-ray diffraction (XRD) and increasing interest in the synthesis of nano crystalline metal
FE-SEM, respectively. In the castables new phases such as oxides.8,9 It is well-known fact that the morphology and
mullite, a-alumina were formed at the expense of bauxite and particle size distribution are important factors that influence
silica. Solid solution of mullite formed at high temperature acts physico-chemical properties of materials, yet the lack of suit-
as a bonding phase and is accounted for high HMOR, CMOR, able, consistent, and low cost methods of preparation have
and CCS values. These excellent properties of such castables limited their ability to control these factors during synthesis.
may enable their uses in various applications such as refractory In the past few decades, there was a great interest in improv-
lining for fabrication of steel, aluminium, copper, glass, ing the quality of refractory castable through decreasing
cement, chemicals, and ceramics. the cement content, i.e., 4%–8% in LCC or 1%–3% in
ultra-low cement castables (ULCC) or <1% in zero cement
I. Introduction castables (ZCC). Decreasing the CaO content in these casta-
bles eliminates the disadvantages of the conventional casta-
appeared in the refractory industry that triggered new pros- (A) Initial Setting of HAC: This early period in the
pects for castable producers. The additives comprise ultra-fine hydration and strengthening of cement is referred to as
particles, for example silica, alumina, mullite, spinel, and “initial setting” of cement. The initial setting time was mea-
carborundum. In previous investigations12 lower amount of sured by taking 50 g of HAC mixed with the percentage of
cement in conjunction with fine grain-sized materials having water required for normal consistency. In Vicat’s apparatus a
high surface area and dispersing aids were used. Those materi- needle is allowed to penetrate through the cement block pre-
als were silica fume, Al2O3, Cr2O3, ZrO2, TiO2, SiC, clay pared. In the initial stage, a thicker needle is allowed to
mineral, and carbon. It was concluded that the presence of pierce through the test block. This procedure is repeated
such materials promote refractory properties such as volume until the paste starts losing its plasticity, and the penetration
stability and well balanced physical and hot mechanical is limited only to 5–7 mm depth. This duration of setting it
properties. The presence of zircon with fine alumina as a filler counted and termed as initial setting time.1
tends to form mullite/ZrO2 composite after firing. The forma- (B) Final Setting of HAC: The cement will be consid-
tion of mullite and zirconia is of vital importance as these ered finally set when, upon, lowering the attachment gently
bonding phases exhibit high refractoriness, low creep rate, low cover the surface of the test block, the thinner needle makes
thermal expansion, good chemical and thermal stability, and an impression. The duration of this process is considered
good toughness and strength.13 Nowadays, low and ULCC final setting time.1
containing SiC have gained wide acceptance as materials for
hot metal and slag runners. Silicon carbide is being introduced
to enhance the thermo mechanical properties of the composite (5) Castable Formulation
under operating conditions. Some authors studied high- Low cement refractory castables are generally prepared using
temperature strength and thermal shock of ultra and LCC in approximately 5 wt% HAC cements and refractory grade
relation to their microstructure feature. These castables have high alumina containing bauxite with certain superfine addi-
been composed of SiC-containing bauxite. They found that tives. The calcined Chinese bauxite that was used in castable
castable specimens had significantly improved in HMOR formulation in this study contained approximately 88.60%,
studied at 1400°C.14 However, the problem of SiC oxidation 4.78%, 1.58%, 4.0%, 0.26%, 0.08%, and 0.70% by weight
limits its presence in refractory castables. The aim of the Al2O3, SiO2, Fe2O3, TiO2, CaO, Na2O, and others, respec-
present work was to synthesize a novel co-melt precursor tively, as specified by the supplier. In the bauxite matrix
HAC and their implementation in bauxite containing castable small additions of reactive alumina, micro-fine silica, and SiC
to investigate the sintering effect of SiC, silica fume, and powder was made. The formulation Tables I and II show the
reactive alumina. detailed composition with their specific names. The effect of
these additions on thermo-mechanical and structural proper-
II. Experimental Procedure ties was also observed. In the first step for cement castable
formulation, calcined bauxite was oven dried, crushed, and
(1) Material Characterization grounded for grading into different sizes in a planetary ball
The starting A.R. grade raw materials aluminum nitrate, mill. The jar and grinding media were of titanium-coated
calcium nitrate, silica fumes, silicon carbide powder, and stainless steel material. At one time 250 g of calcined bauxite
reactive alumina were supplied by Loba Chemie Pvt. Ltd., material was taken in a jar and grounded in a high-energy
Mumbai India. planetary ball mill for 45 min at 600 rpm. Similarly, it was
processed to complete the grinding of complete material. The
ground material was then kept in various selected sieves and
(2) HAC Powder Preparation set up on the motorized vibro sieving equipment for grading.
Hydrated aluminum nitrate Al(NO3)39H2O and hydrated After separation of different graded bauxite, experiments
calcium nitrate Ca(NO3)24H2O were weighed and mixed on were performed further. Figure 1 is the pictorial representa-
a laboratory scale to give the equivalent molar ratio of 7:3 tion of particle size analysis done with the help of standard
and 8:2 Al2O3/CaO, which were denoted as HAC70 and Tyler mesh. The particle size distribution of the fine fraction
HAC80 HAC samples, respectively. The mixtures were is generally a representation of the flow characteristics.
heated in a porcelain dish on a hot plate at about 250°C The particle size distribution has an important role in the
until it melted completely, and were then air quenched. The
resulting solids were heated at 500°C for 3 h. Samples of the
HAC powders obtained were finally calcined at 800°C, Table I. Batch Composition with HAC70
900°C, and 1000°C in a platinum dish using a SiC muffle
HAC70 Chinese SiC Reactive Microfine
furnace for soaking period of 10 min. After soaking, the
Sample (wt%) Bauxite (wt%) (wt%) Al2O3 (wt%) SiO2 (wt%)
powders were quenched and stored in a desiccator.
R1 5 80 5 5 5
R2 5 80 5 6 4
(3) Consistency of Cement R3 5 80 5 7 3
Consistency is a measure of plasticity of a cement paste. It R4 5 80 5 8 2
refers to the degree of wetness exhibited by a freshly mixed R5 5 80 5 9 1
concrete, mortar, or neat cement ground whose workability R6 5 80 5 10 0
is considered acceptable for the purpose at hand. It is mea-
sured as the amount of water required as to a particular wt
% of dry cement which permits the Vicat’s plunger of Table II. Batch Composition with HAC80
10 mm diameter to penetrate to a point 5–7 mm from the
bottom of Vicat’s mold with gauging time 3–5 min. Calcined HAC80 Chinese SiC Reactive Microfine
Sample (wt%) Bauxite (wt%) (wt%) Al2O3 (wt%) SiO2 (wt%)
HAC powder having 70% alumina and 80% alumina
content samples were tested for their consistency by Vicat’s S1 5 80 5 5 5
apparatus.1 S2 5 80 5 6 4
S3 5 80 5 7 3
S4 5 80 5 8 2
(4) Setting Time Test S5 5 80 5 9 1
Calcined powders were tested for initial and final setting by S6 5 80 5 10 0
Vicat’s apparatus according to ASTM C403.
Low Temperature Synthesis of HAC 3
properties of refractory castable. Incorrect particle size distri- the phase formation and changes over the temperature range
bution may cause militancy or the excess water requirement and variation in compositions were carried out.
by the castables. The trials of aggregate proportions were
taken in a 1000 cm3 flask filled up to 250 cm3 and vibrated
for 30 s and the packing density calculations were carried (7) Cold Crushing Strength of HAC and LCC Formulated
out for each trial. Aggregates having highest packing densi- The CCS is the capacity of a material to withstand axially
ties were chosen for further analysis. In the next step, batches directed pushing forces. By definition, the compressive
were prepared by taking different grades of materials and strength of a material is that value of uniaxial compressive
additives in the proper proportion and are summarized in strength reached when the material fails completely. CCS of
Tables I and II. The materials were dry mixed in a plastic cement was measured as the compressive strength of a
container for 10 min by spatula and were then used in sam- 50 mm cement cube made of pure HAC (without any aggre-
ple preparation. Generally, ultra and LCCs require less than gates). These samples were tested for compressive strength
5 wt% of water to achieve the desired rheology; therefore, after 6, 24, and 48 h, respectively, according to (ASTM
water was added in two steps. The casting was done by C1194-03).
adding first two-thirds proportion of water at a time. Then,
one-third of water was added slowly to get a homogenous
mixing. The wet mixing was performed for up to 5–6 min to (8) Cold and Hot Modulus of Rupture (CMOR and
achieve proper flow. Immediately after doing wet mixing, the HMOR) of LCC Formulated
castable mix was filled into a rectangular bar shape mold Cold and hot modulus of rupture (CMOR and HMOR)
(152 mm 9 25 mm 9 25 mm) made of hard steel. The mold measurements were carried out under three-point bending
was placed on the vibrating table filled with the wet mixed tests (ASTM C133-97for CMOR and ASTM C583-10for
castable and the mixes were vibrated for 10 min, showing HMOR) using 152 mm 9 25 mm 9 25 mm samples. CMOR
better compactness. For each composition, several samples and HMOR were calculated using the following formulae:
were prepared for laboratory testing. The samples were cured
in a moisture-saturated environment (95% RH) in a humid- 3PL
ity chamber at room temperature for different time periods. MOR ¼
2db2
For firing the samples, they were first oven dried at 110°C
for 24 h. The test samples were fired at 1300°C–1550°C with where, MOR = modulus of rupture (MPa), P = maximum
a variation of 5°C in an electric SiC heating element applied at rupture (N), L = span between supports (mm),
furnace with maximum soaking time of 3 h. The cured sam- b = breadth or width of specimen (mm), and d = depth of spec-
ples as well as the fired samples were tested for bulk density, imen (mm). CMOR tests were conducted at room temperature
apparent porosity, HMOR, CMOR, and cold crushing using universal materials test equipment (Model 810, MTS
strength (CCS). These samples were also analyzed by XRD System, Eden Prairie, MN) for samples pre fired at 1300°C,
for phases present and by FE-SEM for their morphological 1350°C, 1400°C, 1450°C, 1500°C, and 1550°C for 3 h. HMOR
behavior. measurements were made at 1400°C, 1500°C, and 1600°C
using Netzsch 414/3 HMOR Equipment (Netzsch, Selb, Ger-
many) for samples pre fired at 1550°C for 3 h, cooled at room
(6) Phase Identification temperature and then reheated for testing.
X-ray diffraction patterns were observed using a Rigaku
portable XRD machine (Rigaku, Tokyo, Japan). Calcined
III. Results and Discussion
powdered samples were analyzed by XRD for phases present
in the fired sample. Phase identification analysis was carried (1) Evolution of Phases by X-Ray Diffraction
out by comparing the respective powder XRD patterns with Figures 2 and 3 are the XRD patterns of samples calcined in
the standard database stated by JCPDF. Detailed analyses of the range 800°C–1000°C with 70% and 80% alumina to be
4 Journal of the American Ceramic Society—Kumar et al.
HAC80 samples have more compressive strength than the (4) FE-SEM of HAC
HAC70 sample. The higher strength of HAC80 is ascribed to Figures 6(a) and (b) represents FE-SEM of networks of
the presence of larger amounts of CA and CA2. It is well- well-crystallized interlocking hexagonal plates of CA matrix
known that CA and C12A7 react significantly in the early of HAC70 and HAC80 fired at 1000°C, respectively. In both
ages of hydration, and the hydration of C12A7 is very figures a, b, c, d represent different magnifications of same
exothermic, so the formation of stable hydrates generally samples. The monoclinic phases of CA2 and cubic C12A7
occurs sooner. Although CA2 is known to react slowly with can be observed in both figures. High amount of CA and
water in the early stages of hydration, its presence along with CA2 is proposed to enhance the refractory properties. C12A7
other phases results in an overall faster hydration rate as the phase is responsible for high bonding and quick hydration
heat of hydration resulting from the hydration of CA acti- and small cubic crystals of this phase are evident in the
vates CA2 and makes it react relatively faster with water most homogeneous distribution of CA phase matrix.
than it would do alone, but not more than C12A7. CCS data Figure 6(c) is the EDS representation of both HAC70 and
of the sintered ceramic bodies exhibit high strength of both HAC80, which is a further confirmation of the exact elemen-
samples. This is due to the formation of the ceramic bond tal composition and it also a depiction of pure phases
and absence of any impure phase.16–19 obtained.
(a)
a b
(c)
c d
(b)
a b
c d
Fig. 6. (a) FE-SEM of HAC70; (b) FE-SEM of HAC80; (c) EDS pattern of HAC70 and HAC80.
6 Journal of the American Ceramic Society—Kumar et al.
(5) XRD Patterns of Castables Fired at 1550°C for 3 h more mullite formation. These peaks were identified by stan-
Figures 7 and 8 portray XRD patterns of castables fired at dard JCPDS cards numbered 46-1212, 15-0776 for Fig. 6
1550°C for 3 h, here R1 to R6 and S1 to S6 indicate casta- and 46-1212, 06-0258 for Fig. 8. Minority peaks of silicon
bles formulated with HAC70 and HAC80, respectively. The carbide in both figures were identified and matched with
corundum phase appeared as a major component due to the JCPDS card number 42-1091.
transformation of bauxite minerals to corundum. The second
major phase was detected as mullite, which resulted due to
the addition of microsilica and reactive alumina in all the (6) Bulk Density and Apparent Porosity
castable samples. In both castable series R and S, as we Table III shows the values corresponding to bulk density,
proceed from high to low batch number, the peak intensities apparent porosity, and thermal shock of R3 and S3 series
of mullite increase this is accounted for increasing micro fine castables sintered at varying temperature range. Only these
silica content. Although SiC phase was added as a minor series was selected as homologous to their other thermo-
ingredients, but it fails to show its intensity peak in all the mechanical superiority. In the castables prepared, highest
plots, so it could be extracted from the fact that at high bulk densities achieved are 2.91 and 2.93 g/cm3 for the com-
temperatures its oxidation would have occurred forming positions R3 and S3, respectively. This can be attributed to
silica and finally reacting with aluminus content to result in the use of higher percentages of alumina and micro-fine silica
in the aggregate which resulted in most mullite solid solution
formations. Apparent Porosity was measured to be in the
range of 8%–11% in all the castables prepared.
Table III. Bulk Density, Apparent Porosity, and Thermal Shock Cycles of Prepared Castables
Temperature (°C)
Test method Sample name 1300°C 1350°C 1400°C 1450°C 1500°C 1550°C
(a) (b)
(a) (b)
(a)
a b
(b)
c d
Fig. 12. (a) FE-SEM of S3 castable prepared with HAC80; (b) EDS pattern of S3 castable.
2
castable and reinforced the matrix confirming the high P. C. Hewlett, “Calcium Aluminate Cements”; pp. 709–71 Lea’s Chemistry
strength and refractory properties. On the other hand, the of Cement and Concrete, 4th edition, Edited by P. C. Hewlett. Butterworth
Heinmann, New Delhi, India, 2004.
impurities that were present in Chinese bauxite may have 3
S. R. Klaus, J. Neubauer, and F. Goetz-Neunhoeffer, “Hydration Kinetics
formed a low melting glassy phases, which tend to decrease of CA2 and CA-Investigations Performed on a Synthetic Calcium Aluminate
the refractory properties and their microstructure may Cement,” Cem. Concr. Res., 43, 62–9 (2013).
4
appear as partially diffused structure in some regions. This K. L. Scrivener, J. L. Cabiron, and R. Letourneux, “High-Performance
Concretes From Calcium Aluminate Cements,” Cem. Concr. Res., 29, 1215–23
property is more evident when we zoom into the microstruc- (1999).
tures while moving a to d in each plot. The EDS pattern of 5
H. G. Midgley, “High alumina cement in construction- a future based on
S3 castable is portrayed in Fig. 12(b), which is in correspon- experience”; pp. 1–13 in Calcium Aluminate Cement, Edited by R. J. Mangab-
dence with exact stoichiometric composition of the hai. Chapman & Hall, London, 1990.
6
S. P. Jiang, J. C. Mutin, and A. Nonat, “Studies on Mechanism and Physi-
formulated block. The influence of trace impurities presented cochemical Parameters at the Origin of the Cement Setting: I. The Fundamen-
during the processing of ceramics is well-known and tal Processes Involved During the Cement Setting,” Cem. Concr. Res., 25 [4]
exploited in the control of densification, grain growth, and 779–89 (1995).
7
morphology during sintering.20 The ability to control micro- A. Capmas, D. Menetrier-Sorrentino, and D. Damidot, “Effect of Temper-
ature on Setting Times of Calcium Aluminate Cements”; pp. 65–80 in Calcium
structure is important for achieving desired properties. Aluminate Cements, Edited by R. J. Mangabhai. Chapman & Hall, London,
1990.
8
R. N. Das, A. Pathak, and P. Pramanik, “Low-Temperature Preparation
IV. Conclusion of Nanocrystalline Lead Zirconate Titanate and Lead Lanthanum Zirconate
Titanate Powders Using Triethanolamine,” J. Am. Ceram. Soc., 81 [12] 3357–
Novel process synthesized HACs were successfully prepared 60 (1998).
with very small crystallite size ranging 20–35 nm. Cementing 9
B. Myhre, “Strength Development of Bauxite-Based Ultralow-Cement
behavior promises a new era of industrial evolution through Castables,” Am. Ceram. Soc. Bull., 73 [5] 68–73 (1994).
10
commercial implementation. Desired cementing phases having M. F. M. Zawrah and N. M. Khalil, “Effect of Mullite Formation on
Properties of Refractory Castables,” Ceram. Int., 27 [6] 689–94 (2001).
high refractoriness similar to CA, CA2, C12A7 were formed at 11
M. A. Serry, M. F. Zawrah, and N. M. Khalil, “Bauxite-Based Low and
a very low temperature. One advantage of this process is to Ultra Low Cement Castables,” Br. Ceram. Trans., 101 [4] 165–8 (2002).
12
exclude silicate phases having low eutectic temperatures in G. Maczura, J. Kopanda, and F. Rohr, “Calcium Aluminate Cements for
such cements. These silicate phases decrease the refractoriness Emerging Castable Technology”; pp. 285–304, Advances in Ceramics, 13, New
Development in Monolithic Refractories, Edited by R. E. Fisher. Am. Ceram.
and the cement cannot be used at higher temperatures. The Soc., Columbus, OH, 1985.
XRD patterns of castables containing Bauxite and SiC show 13
E. Y. Piviniski, “New Refractory Concrete and Bonding Systems,”
new phases formed such as corundum and mullite, which Refract. Ind. Ceram., 39 [314] 91–9 (1998).
14
were responsible for superior thermo-mechanical and physical W. Kronert and U. Schumacher, “Use of Low-Cement and Ultra-Low
Cement Refractory Castables in Iron and Steel Industry Furnace Prospect,”
properties. Castable samples prepared by HAC having 80% Interceram, Aachen Proc., 3 [8] 12–8 (1989).
alumina have better physical and cementing properties than 15
N. Richard, N. Lequeux, and P. Boch, “An X-ray Absorption Study of
the HAC having 70% alumina. Needle like mullite formed at Phases Formed in High-Alumina Cements,” Adv. Cem. Res., 7 [28] 159–69 (1995).
16
high temperature acts as a bonding phase which was responsi- V. K. Singh and U. K. Mandal, “Kinetics Study of the Thermal Synthesis
of Calcium Aluminate Above 1400°C,” Trans. J. Br. Ceram. Soc., 81 [4] 112–3
ble to increase the CMOR. FE-SEM represents dense micro- (1982).
structure of all the samples and glassy phase appearance is 17
V. K. Singh, M. M. Ali, and U. K. Mandal, “Formation Kinetics of
due to the impurities present in bauxite. Calcium Aluminates,” J. Am. Ceram. Soc., 73 [4] 872–6 (1990).
18
V. K. Singh and M. M. Ali, “Formation Kinetics of High Alumina
Cements Phases,” J. Br. Ceram. Soc., 79 [4] 112–4 (1980).
19
References M. F. Zawrah and N. M. Khalil, “Utilization of Egyptian Industrial
Waste Materials in Manufacture of Refractory Cement,” Br. Ceram. Trans.,
1
V. Kumar, V. K. Singh, A. Srivastava, and G. N. Agrawal, “Low Temper- 101 [5] 71–4 (2002).
20
ature Synthesis of High Alumina Cements by Gel-Trapped Co-Precipitation S. Mukhopadhyaya and P. K. Das Poddar, “Effect of Preformed and In
Process and Their Implementation as Castables,” J. Am. Ceram. Soc., 95 [12] Situ Spinels on Microstructure and Properties of a Low Cement Refractory
3769–75 (2012). Castable,” Ceram. Int., 30, 369–80 (2004). h