See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/281996951

“Reaction of Orthophosphoric Acid with Several Forms of Aluminum Oxide,”

Article  in  American Ceramic Society Bulletin · July 1980

CITATIONS READS

36 3,582

2 authors, including:

John W. Halloran
University of Michigan
277 PUBLICATIONS   9,106 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

DARPA Metamaterials Project View project

All content following this page was uploaded by John W. Halloran on 25 September 2017.

The user has requested enhancement of the downloaded file.

Reaction of Orthophosphoric Acid with
Several Forms of Aluminum Oxide
FRANCISCO J. GONZALEZ* The reaction of orthophosphoric acid with
Pennsylvania State University, University Park, Pa. 16802 several types of Bayer-process and am­
monium alum-derived aluminas was studied
and at temperatures up to 500°C. The reaction
JOHN W. HALLORAN* products consisted of a mixture of aluminum
Case Western Reserve University, Cleveland, Ohio 44106 phosphates that vary depending on the type
of alumina, concentration, reaction time, and
temperature. For all temperatures, the ex­
tent of the reaction—as given by the fraction
Phosphate bonding originates from the reaction of orthophos- of alumina reacted—and the relative yields
phoric acid or acidic phosphates with oxides or siliceous materi­ of each aluminum phosphate were shown to
als at ambient or high temperatures. Amphoteric oxides are 1-4

be related to the total amount of surface area

most favored because they react at a controllable rate. The process
of phosphate bonding in high alumina refractories involves the of the powder per mole of Ρ205 in the sys­
formation of acid phosphates, which on subsequent heating in the tem. Condensed phosphates, obtained by
presence of excess alumina react with the oxide, resulting in the condensation of previously formed acid
formation of crystalline or amorphous aluminum phosphates. A phosphates, were proved to be still capable
stable bond is thus developed at curing temperatures as low as
of reacting with excess alumina at higher
300°C, and the bond matures during the burning process, usually
conducted in situ. The final bond is almost certainly A1P0 in any of temperatures. Alum-derived metastable
γ-Α l 2 0 3 , reacted quickly at low tempera­
4

its various polymorphic forms. Phosphate-bonded high alumina

5-7

products, available in the brick and specialties forms, have unusual tures, yielding crystalline AIP04. Alum-
qualities—such as high abrasion resistance, high load-bearing derived a-AI203 powders reacted slowly,
capability, and dimensional stability—that make them preferred yielding amorphous products.
materials for a number of applications. 8-10

Literature data on the reaction of orthophosphoric acid with

alumina are numerous and contradictory. Aluminum is known to be
capable of forming different solid phases with phosphate depending
on the initial concentration of the two reactants and the pH. 11-13
phoric acid-alumina mixes because of the relative inertness of
Because of the similarity of their chemical properties, the lack of alumina. Hence, several papers dealing with the thermal transfor­
clarity in their nomenclature, and the complex procedure for their mations of monoaluminum phosphate binders have appeared. 20-21

preparation and separation, it is difficult to identify the different One problem that has been the subject of study is the premature
phases. hardening of phosphate-bonded high alumina refractory compo­
Sheets et al. concluded that phosphate-bonding in high alumina
3
sitions. Lyons et al. suggest that premature hardening occurs as a
22

refractories occurs because of covalent polymerization of aluminum result of room-temperature reaction between the Al(OH) layers of 3

phosphates, which proceeds in several stages upon heating. clay minerals and H P0 to form, initially, A1(H P0 ) and, finally,
3 4 2 4 3

Klyucharev and Skoblo claim that ambient temperature setting is

14
precipitate A1P0 -3H 0. More recently, Lyon et al. and Eti and
4 2
23

due to the formation of variscite (A1P0 -2H 0), whereas setting 4 2 Hall have suggested that the following reactions are responsible
24

during heating is due to the formation of an amorphous product. for the early setting of high alumina mixes:
This theory is in agreement with solubility studies in the system
AI2 O3 -P2 O5 -H2 O under the condition of excess alumina. 15-17

Kingery, on the other hand, attributes setting and hardening in

1

aluminum hydroxide-phosphoric acid mixes to the formation of

monosubstituted aluminum phosphate, A1(H P04)3, which upon 2
Prendergast suggests that Α1Η (Ρ0 ) ·χΗ20 is responsible for the
25
3 4 2

heating yields aluminum metaphosphate, A1(P03)3. Kopeikin et premature setting according to the following scheme:
al. and Yutina et al. maintain that the ability of mixes containing
18 19

orthophosphoric acid and alumina to acquire strength and moisture

resistance when heated to 270°C is not due to the formation of
anhydrous A1P0 in any of its various crystalline forms, but rather
4

to the formation of an amorphous phosphate. He also asserts that the presence of chemically active ions such as
Many investigators have suggested that phosphoric acid-
1 4 1 8- 1 9
Fe3+, Al3+, Ca2+, Mg2+, and β-Α1 0 play an important role in
2 3

alumina reactions lead to the formation of phases similar to those premature setting.
produced by the phosphoric acid-aluminum hydroxide reactions. Recent review papers by Cassidy and Chvatal point out that in
26 27

The usual procedure has been to assume that whatever changes real refractory systems an excess of some reactive phase such as
occur in certain aluminophosphate binders or binders with small alumina, silica, zirconia, etc., is always present. Thus bonding
additions of aluminum hydroxide will also occur in the orthophos-
14
occurs in two ways: by chemical bonding to the aggregate and by the
formation of acidic or condensed phosphates, which at certain
temperatures may decompose into polymeric phosphates which are
*Member, the American Ceramic Society. mostly amorphous. The work of O’Hara et al. shows this feature
28

very clearly. The drop in strength observed in some phosphate-

bonded high alumina refractories has been attributed by these inves­
Presented at the 81 st Annual Meeting of the American Ceramic Society, Cincinnati, tigators to the recrystallization of the amorphous phases that takes
Ohio, May 1. 1979 (No. 6-JII1-79).
Received December 12, 1979; revised copy received March 24. 1980; approved place at still higher temperatures, accompanied by liberation of
April 1, 1980. p2o5.
Vol. 59, No. 7 (1980) 727
 with the aim of achieving better understanding of the mechanism of
phosphate bonding in high alumina compositions. Attention is cen­
tered on the reactions that occur at low to moderate temperatures
since this is the range of normal curing conditions.

Experimental
Approximately 250 g of alumina-orthophosphoric acid mixes
corresponding to the bulk compositions shown in Table I were
prepared by thoroughly mixing reagent grade orthophosphoric
acid* (85% H P04) and one of several Bayer-process calcined
3

aluminast or one of two different high purity aluminas* obtained by

decomposition of ammonium alum. Typical properties of the
aluminas are listed in Table II.
Surface area determinations were made by single-point BET
dynamic gas adsorption§ using nitrogen as the adsorbate. Thermal
analyses on fresh mixes were conducted in a differential thermal
analyzed and a thermobalance.** All runs were done using air as the
carrier gas and heating rates between 5° and 20°C/min.
Standard aluminum phosphates were synthesized from reagent
grade Al(OH) and H P0 according to methods described
3 3 4

elsewhere. Calibration curves for the quantitative determina­

1 3, 3 1 - 32

tion of reaction products and unreacted alumina were obtained by

mixing known amounts of standard aluminum phosphates with
α-Α1 0 and Zr0 as an internal standard. A diffractometer* with
2 3 2
1

Fig. 1 . Effect of specific surface area and AI2 O3 /P2 O5 molar

Ni-filtered CuKa radiation was used to record diffraction patterns at
ratio (R) on (A) fraction of alumina reacted and (B) yields of
a scanning speed of 1/4 to 1/2 °2θ/min. The characteristic XRD peaks
AI(H P0 ) and B-AIP0 in mixes of Bayer alumina and or-
2 4 3 4
used for each aluminum phosphate are given in Table III.
thophosphoric acid reacted 1 0 days at 150°C.
Small portions of the alumina-phosphoric acid mixes were placed
in clean porcelain crucibles and reacted at temperatures up to 500°C
The use of different starting materials has a pronounced effect in for various periods of time. Periodic inspection of the reacted mixes
the phases formed by the reaction of orthophosphoric acid with was performed by quantitative X-ray diffraction analysis (QXDA)
alumina. Since it is not unreasonable to expect that in many
29-30
and acidimetry. The extent of the reaction was calculated from the
cases equilibrium is not achieved, kinetic considerations such as the relative amounts of crystalline aluminum phosphates determined by
reactivity of the materials, concentration, and curing time and QXDA in relation to the maximum possible conversion—as given
temperature play an important role in determining the bonding by direct theoretical calculations—for each composition. No pre­
phases. Highly reactive materials such as aluminum hydroxide, caution was taken to minimize P 0 loss. 2 5

transition aluminas, and calcined aluminas of high surface area will

react quickly even at room temperature. On the other hand, rela­ Results and Discussion
tively inert materials such as calcined aluminas of low surface area Table IV lists the results obtained by QXDA after reaction of the
and sintered and fused aluminas will react slowly, and higher curing alumina-orthophosphoric acid mixes at several temperatures for 1 0
temperatures could be required to obtain a stable bond, which may days. The crystalline reaction products consisted of a mixture of
be physically and chemically different from the bond developed aluminum orthophosphate in the berlinite form (B-AIPO4 ), the
with more reactive forms of alumina. tridymite form (T-AIPO4 ), and the cristobalite form (C-AIPO4 ),
The objective of the work reported herein is to show the effect of monoaluminum phosphate, A1(H P04)3, aluminum dihydrogen 2

starting materials on the reaction of orthosphoric acid with alumina tripolyphosphate A 1Η Ρ Ο ·2Η Ο, and the B-type of A1(P03)3. At
2 3 10 2

temperatures below 150°C the reaction pro­

ceeded slowly, and only traces of
A1(H P0 ) and A1H (P04)2· 3H20 were de­
2 4 3 3

tected.
The reaction of orthophosphoric acid
with alumina, like other reactions involving
solid particles, occurs at the alumina sur­
face. Thus the kinetics of the reaction is
affected by the total surface area

*J. T. Baker Chemical Co., Phillipsburg, N.J.

tAluminum Co. of America, Pittsburgh, Pa.
tUnion Carbide Corp., Linde Div., Indianapolis, Ind.
§Quantachrome Corp., Syosset, N.J.
j|Harrop Labs, Columbus, Ohio.
**Robert L. Stone Co., Austin, Texas.
ttPhilips Electronic Instruments, Inc., Mahwah, N.J.

728 Ceramic Bulletin

available for the reaction to take place. The fraction of acid reacted
at a given temperature and time will increase as the quantity of
alumina increases and as the size of individual alumina particles Fig. 2. Effect of relative activity on fraction of alumina reacted
decreases. Commercial powders consist of irregular particles of a in mixes of Bayer alumina and orthophosphoric acid reacted at
wide variety of sizes; hence the specific surface area of the powder, temperatures indicated for 10 days.
rather than its average particle size, should be used for the analysis
of macroscopic observations. This is particularly important in
solid-fluid reactions, where the reactant fluid is capable of wetting reacted and acid concentration on the yields of A1(H P0 ) and2 4 3

the irregular particles just as well as the adsorbate used in the B-A1PO after reaction for 10 days at 150°C are shown in Fig. 1(B).
4

measurement of the surface area. As can be seen, the relative yield of the acid phosphate, A1(H P04)3,
2

Figure 1 (A) shows the effect of the specific surface area of the is higher in mixes prepared with aluminas of low surface area and
Bayer-process aluminas on the fraction of alumina reacted after 10 low A1 0 :P 0 ratio. On the other hand, higher yields of aluminum
2 3 2 5

days at 150°C. As expected, the quantity of alumina reacted in­ orthophosphate, B-A1P04, are observed in high surface area mixes
creases with increasing surface area and increasing A1 0 :P 0 2 3 2 5 or where the A1 0 :P 0 ratio is high.
2 3 2 5

molar ratio R. The effects of the specific surface area of the alumina Figure 2 shows the combined effects of the specific surface area

Vol. 59, No. 7 (1980) 729

 of the Bayer-alumina powder and the Αl2 O :Ρ O 5 molar ratio of the
3 2

mix on the fraction of alumina reacted at several reaction tempera­

tures. The relative activity of the system, SR, is given by the total
surface area of the solid per mole of P O5 in the system. A system of
2

high relative activity can be obtained by using a powder of high

specific surface area or by preparing a mix containing enough
alumina to yield a high A1 0 :P 0 ratio. In contrast, a mix of low
2 3 2 5

relative activity can be prepared by using a coarse alumina and by

introducing sufficient acid into the system to make the resulting
A1 0 :P 0 ratio low.
2 3 2 5

As is shown in Fig. 2, the fraction of alumina reacted increases

with increasing relative activity and temperature. Individual trends
of the Bayer aluminas at 150°C are indicated by broken lines. From
Fig. 3 it can be seen that, for mixes with low relative activity values,
the yield of aluminum phosphates with low ALP atomic ratio (i.e.,
Francisco J. Gonzalez John W. Halloran A1(H P04)3, AlH P Oi0 -2 H O, and A1(P03)3) is relatively high and
2 2 3 2

decreases with increasing relative activity. In contrast, phosphates

of a high A1:P ratio (i.e., the polymorphs of AlP04) are present in
Francisco J. Gonzalez is a Ph.D. candidate at Pennsylvania
mixes of high relative activity.
State University, University Park. He received his B.S. in
The complexity of the reaction of orthophosphoric acid with
chemical engineering from Monterrey Institute of Technology,
alumina is due in part to the large number of possible reaction
Mexico, in 1975 and his M.S. in ceramic engineering from the
products and intermediate steps involved. At the beginning and at
University of Missouri-Rolla in 1978. His main interests are
moderately low temperatures, the reaction can be viewed as a
processing and physical properties of refractories.
dissolution process in which aluminum ions dissolve into the acid.
John W. Halloran is assistant professor of ceramics at
The process involves several steps: transport of reactant to the
Case Western Reserve University, Cleveland, Ohio. He
interface, reaction at the interface, and transport of the products
earned a B.S. in ceramic engineering from the University of
away from the interface. The reaction most certainly yields the
Missouri-Rolla in 1973 and a Ph.D. in ceramics from Mas­
primary soluble phosphate, A1(H P04)3, according to Eq. (1).
sachusetts Institute of Technology in 1977. Before joining 2

However, solutions of monoaluminum phosphate are unstable, and

CWRU this year, Dr. Halloran was an assistant professor at
A1(H P0 ) should be considered an acidic salt of aluminum or­
Pennsylvania State University. His research interests include 2 4 3

thophosphate with two loosely attached molecules of orthophos­

ceramic processing, diffusion in oxides, and refractories.
phoric acid15,25:

Kinetic analysis shows that when the reaction rate is controlled by

31

either the interface reaction rate or by mass transport, the dissolution

process can be represented by

where dna /dt is the molar dissolution rate, K is a constant, φ(t)

involves diffusion coefficients and characteristic thickness for the
transport-controlling case and the specific reaction rate constant for
the interface reaction-controlling case, and SR is the relative activity
of the system. Thus the time necessary to reach saturation through­
out the liquid is inversely proportional to the relative activity of the
system, SR.
Further attack to the alumina grains after saturation leads to the
precipitation of hydrated phosphate compounds:

Fig. 3. Effect of relative activity and reaction temperature on yields of

aluminum phosphates in mixes of Bayer alumina and orthophosphoric
acid reacted for 10 days at (A) 150°, (B) 300°, and (C) 500°C.

730 Ceramic Bulletin

Fig. 4. Thermal transformations of aluminum phosphates.

phosphates detected by QXDA were affected by the type of alumina

and bulk composition. The acidic phosphate, A1(H P04)3, was
2
where X varies from 1 to 3. Under humid conditions
observed in compositions with low A1 0 :P 0 molar ratio and in a
2 3 2 5
A1H (P0 ) *3H20 is the most stable phase. This phosphate is re­
3 4 2
lesser quantity in the mixes prepared with the more reactive AA-2
ported to dehydrate reversibly to A1H (P0 ) *H20 at 90°C and to
32
3 4 2
alumina, which showed larger amounts of aluminum orthophos­
decompose into A1(H P0 ) and A1P0 at 150°C. Precipitation of
2 4 3 4
phate, B-A1P04. These results are not directly comparable to those
A1H (P0 ) 'XH20 is likely to occur slowly since the process in­
3 4 2
obtained with the Bayer aluminas because of the marked differences
volves a nucleation step in addition to diffusion of reactants through
in crystallinity, particle size, and impurity levels. Handling of the
a layer of solid product and chemical reaction. Nevertheless, the
aluminas after calcination (such as milling) may also contribute to
kinetics of the reaction is directly proportional to the relative activity
the difference in the results, since milling has been observed to 36

of the system, SR.

influence powder reactivity and Bayer aluminas are commonly
Formation of hydrated aluminum orthophosphates may occur if
milled after calcination but alum-derived aluminas are not.
excess alumina is present, since A1(H P0 ) is still capable of 2 4 3
The absence of crystalline phosphates in mixes containing AA-1
reacting:
alumina in high A1 0 :P 0 ratios indicates that, in this case, the
2 3 2 5

reaction yields an amorphous product after the initial formation of

A1(H P04)3. In contrast, mixes of identical composition prepared
2

Under dry conditions (such as when temperatures are above 100°C), with AA-2 yield large amounts of C-A1P04. The reaction appears
anhydrous A1P0 is likely to form.
4
to follow a different path because of the high reactivity of metastable
The effect of temperature on isolated aluminum phosphates γ-Α1 03, which is present in large proportions in AA-2 and absent
2

has been reported elsewhere. The most important physico­

1 3 , 2 0, 3 1 - 3 2
in AA-1.
chemical changes for some aluminum phosphates are shown in Fig.
4. It may be concluded that the appearance of poly- and metaphos­ Conclusions
phates at 300° and 500°C in mixes of low relative activity originates 1) The reaction of orthophosphoric acid with aluminum oxide
through condensation reactions from monoaluminum phosphate, yields several different products depending on type of alumina, bulk
since Reaction (6) may not proceed to completion. However, since composition, reaction time, and temperature.
excess alumina is always present, products closer to the alumina 2) The combined effects of specific surface area and alumina
particles are sufficiently high in aluminum content to form A1P04. concentration (relative activity) determine the kind and relative
This result indicates that an increase in the total conversion of A1 0 2 3
(Continued on page 738)
should be expected with increasing temperature because of the
solid-state reaction:

As is shown by Fig. 3, the relative activity of the system also affects

the amount of alumina reacted. The effect of temperature in low
relative activity mixes is much more pronounced at 500°C than at
300° or 150°C.
The exothermic peak between 500° and 700°C in DTA diagrams
(Fig. 5) obtained by rapidly heating fresh mixes containing excess
A1 0 is probably due to this solid state reaction. The DTA peak is
2 3

not reversible upon cooling, and the TG plot (Fig. 5) does not
indicate major change within that region. The DTA curve obtained
from a mechanical mixture of A1 0 and A1(P0 ) in a 1:1 molar 2 3 3 3

ratio shows also an exotherm within the same temperature range,

indicating that the exotherm observed in Fig. 5 is almost certainly
due to Reaction (7). X-ray diffraction patterns obtained from mix­
tures heated to 500° and 700°C in the DTA unit show a decrease in
the intensity of the A1(P0 ) peaks with increasing temperature.
3 3

The study of the reaction of orthophosphoric acid with the highly

reactive ammonium alum aluminas (AA-1 and A A-2) was limited
to temperatures up to 150°C. The results after reaction for 10 days at Fig. 5. DTA and TG curves taken at heating rates indicated for mixes
150°C are listed in Table IV. Once again, the amounts of crystalline of Bayer alumina and orthophosphoric acid.

Vol. 59, No. 7 (1980) 731

 5) Use of quantitative thermal shock data in regression analysis research. The guidance and assistance of the Technical Advisory
and computer modeling or simulation studies. Committee of TRI throughout this work has been beneficial. The
6) Based on the test controls available, especially thermal gra­ assistance of the refractories industry, including various producers
dient, a further enhanced understanding of fracture mechanics cal­ and users, in supplying test samples, assisting in comparative test­
culations related to thermal shock resistance. ing, and offering pertinent comments and criticisms, has been
Continued research with the ribbon test is expected to broaden the extremely helpful throughout this work.
utility of the procedure. Various industrial laboratories have ob­
tained the equipment construction details from the Refractories References
'C. E. Semler, “Concurrent and Future Trends in the Refractories Industry”; pp.
Research Center and are in the process of adding this test capability. 45-59 in the 30th Anniversary Commemorative Issue, Japan Technical Assn. of
The increased number of test units, both in the U.S. and foreign Refractories, Tokyo, July 1977.
2 “Standard Method for Panel Spalling Testing Refractory Brick,” 1979 Annual
countries, will greatly expand the data base for further evaluation of Book of ASTM Standards, Pt. 17, pp. 18-39. American Society for Testing and
the test method. Based on the growing interest in the ribbon test, the Materials, Philadelphia, Pa.
3D. M. Bolla and R. A. Strimple, “Reducing Thermal Damage to BOF Linings
method will almost certainly be considered as a standard test for
During Burn-In, ” Proc. Natl. Open Hearth Basic Oxygen Steel Conf, 58, 113-20
assessing the relative thermal shock resistance of commercial re­ (1975).
fractory products in the near future. Concurrent with that ongoing 4R. A. Strimple, C. R. Beechan, and J. F. Muhlhauser, “Monitoring Brick Thick­

ness and Hot-face Temperatures in EAF Sidewalls,” Proc. Electr. Furn. Conf., 31,
activity, it is expected that the ribbon test will be used to further the 214-18 (1973).
basic understanding of thermal shock mechanisms for a wide range 5J. H. Ainsworth and R. H. Herron, “High Temperature Fracture Energy of

Refractories,” Am. Ceram. Soc. Bull., 55 [7] 655-57, 664 (1976).

of ceramic products, from coarse to fine-grained, heterogeneous 6J. J. Uchno, R. C. Bradt, and D. P. H. Hasselman, “Fracture Surface Energies of
bodies (refractories) to fine-grained, homogeneous bodies, from a Magnesite Refractories,” ibid., pp. 665-68.
7B. Brezny, “Crack Formation in BOF Refractories during Gunning,"ibid., 58[7]
theoretical standpoint. Thus it is hoped that the ribbon test will be an
679-82 (1979).
asset to both the industrial and the scientific communities for their 8G. C. Padgett, British Ceramic Research Assn.; personal communication, April

individual needs relevant to thermal shock evaluation and un­ 1979.

9S. M. Brisbane, Standards Assn. of Australia.; personal communication, May
derstanding. 1979.
I0D. P. H. Hasselman etal., “Fracture Toughness of Refractories,” Pennsylvania

Acknowledgments Science & Engineering Foundation, Final Rept., 1973; 29 pp.

"R.C. Bradt, Pennsylvania State University; personal communication, April 1978.
The authors express grateful appreciation to the Refractories 12Available prism spalling tests: Germany, DIN 51068; Britain, B.S. 1902; Austra­

Institute (TRI), Pittsburgh, Pa., for their ongoing support of this lia, ASR 31, Method 24; U.S., Bethlehem Steel procedure (modified). U

Reactions of Orthophosphoric Acid with Several Forms of Aluminum Oxide

(Continued from page 731)
of the Preparation and Physical Properties of Multivalent Metal Condensed Phosphates:
amounts of phosphates obtained by the reaction at a fixed tempera­ XI,"Bull. Chem. Soc. Jpn., 48 [6] 1830-35 (1975).
14Ya. V. Klyucharev and L. I. Skoblo, “Composition of Products in the Hardening
ture and time. In compositions prepared with low relative activity, of Aluminophosphate Binders in Corundum Refractories, ” Zh. Prikl. Khim., 38 [3]
the products contain a low A1:P atomic ratio. If the relative activity 530-35 (1965).
,5R. F. Jameson and J. E. Salmon, “Aluminum Phosphates: Phase-Diagram and Ion
of the mix is high, the products consist of high Al/P phases.
Exchange Studies of the System Aluminum Oxide-Phosphoric Oxide-Water at 25°C, ”
3) Condensed phosphates, such as A1(P03)3, form by condensa­ J. Chem. Soc., 1954, pp. 4013-17.
tion of acid phosphates and are still capable of reacting with alumina 16J. C. Bosheer, F. A. Lenfesty, and J. F. Anderson, Jr., “Solubility in the System

Aluminum Phosphate-Phosphoric Acid-Water,”./. Am. Chem. Soc., 76[23] 5951-56

in the solid state. (1958).
4) Aluminas derived from ammonium alum behaved differently I7R. Martin, C. Duc-Mauge, and H. Guerin, “Aluminum Phosphates: Equilibrium

Diagram for the System Α120.}-Ρ20-,-Η20 at 60°," Bull. Soc. Chim. Fr., 1960, pp.
from Bayer aluminas. This difference is believed to be due to the 851-56.
differences in crystallinity, particle size, impurity levels, and possi­ 18 V. A. Kopeikin, A. I. Kudryashova, L. N. Kuz’minskaya, I. L. Rashkovan, and I.

bly to milling. V. Tananaev, “Formation of the Amorphous Phase during Cementation of Materials
with Aluminophosphate Binder,"Izv. Akad. NaukSSSR, Neorg. Mater., 3[4] 737-39
5) Compositions containing large quantities of γ-Α1 0 are very 2 3 (1967).
19A. S. Yutina, Z. D. Zhukova, and S. V. Lysak, “Reaction of Phosphoric Acid
reactive and yield crystalline C-A1P04. Compositions made with
with Some Forms of Alumina,” ibid., 2 [11] 2020-28 (1966).
the less reactive high purity α:-Α1 0 yield amorphous products if
2 3 20A. Boulle and F. B. D’Yvoire, “Thermal Evolution of Aluminum Phosphates of

the alumina is present in large quantities. the Group P20.-,/Al203 = 3,” Compt. Rend., 245, 531-34 (1957).
21R. N. Rickies, “The Effect of Heat on Inorganic Cements,” J. Appl. Chem.

Acknowledgment (London), 15 [2] 74-77 (1965).

22J. W. Lyons, G. J. McEwan, and C. D. Siebenthal, “Phosphoric Acid in Soil

F. J. Gonzalez wishes to thank Resco Products, Inc., for their Stabilization: II,” Natl. Acad. Sci.—Natl. Res. Council, Publ. No. 964, 5-24(1962).
23 J. E. Lyon, T. U. Fox, and J. W. Lyons, “An Inhibited Phosphoric Acid for Use in
financial support in the form of a fellowship. High-Alumina Refractories,” Am. Ceram. Soc. Bull., 45 [7] 661-65 (1966).
24E. Eti and W. B. Hall, “Control of Premature Hardening in Phosphate-Bonded
References High-Alumina Refracoty Compositions,” ibid., 50 [7] 604-606 (1971).
1 W. D. Kingery, “Fundamental Study of Phosphate Bonding in Refractories: I,"J. 251. D. Prendergast, “Premature Set of Phosphate-Bonded Refractory Mixes,”

Am. Ceram. Soc., 33 [8] 239-41 (1950). presented at the Fall Meeting of the Refractories Division of the American Ceramic
2S. L. Golynko-Vol’fson and L. G. Sudakas, “Some Principles of the Appearance of Society and NICE, Bedford, Pa., October 9, 1976; for abstract see ibid., 55 [8] 740
Binding Properties in Phosphate Systems, "J. Appl. Chem. USSR (Engl. Transl.), 38 (1976).
26 J. E. Cassidy, “A Review of Phosphate-Bonded Refractories,” in High Tempera­
[7] 1441-45 (1965).
3H. D. Sheets, J. J. Bulloff, and W. H. Duckworth, “Phosphate Bonding of ture Chemistry of Inorganic and Ceramic Materials, ” Spec. Publ. No. 30. Edited by F.
Refractory Compositions,” Refract. J., 34 [9] 402-406 (1958). P. Glasser and P. E. Potter. Chemical Society, London, 1976.
4K. Fisher, “Chemical Bonds for Refractory Materials,” Proc. Br. Ceram. Soc., 27T. Chvatal, “The Position of Refractory Phosphate Bonding Today,"Sprechsaal

12, 51-64 (1969). Keram., Glas, Baustoffe, 108 [19-20] 576, 578, 580, 582-89 (1975).
50. W. Florke, “The Structure of A1P0 28M. J. O'Hara, J. J. Duga, and H. D. Sheets, Jr., “Studies in Phosphate Bonding,”
4 and Si02”; pp. 13-27 in Science of
Ceramics, Vol. 3. Edited by G. H. Stewart, Academic Press, New York, 1967. Am. Ceram. Soc. Bull., 51 [7] 590-95 (1972).
6F. A. Hummel, “Properties of Some Substances Isostructural with Silica, "J. Am. 29G. D. Salmanov andG. N. Aleksandrova, “Highly Refractory Concrete Based on

Ceram. Soc., 32 [10] 320-26 (1949). an Aluminophosphate Binder,” Izv. Akad. NaukSSSR, Neorg. Mater., 5 [1] 148-51
7W. R. Beck, “Crystallographic Inversions of the Aluminum Orthophosphate (1969).
30A. A. Chistayakova, V. A. Sivkina, V. I. Sadkov, and A. P. Khashkovskaya,
Polymorphs and Their Relation to Those of Silica,” ibid., [4] 147-51.
8W. H. Gitzen, L. D. Hart, and G. MacZura, “Phosphate-Bonded Alumina Casta­ “Effect of Corundum on the Phase Composition of Products of Heating in the Al20;r
bles: Some Properties and Applications,” Am. Ceram. Soc. Bull., 35 [6] 217-23 P20.-,-H20 System,” ibid., [10] 1738-41.
31F. J. Gonzalez, “Phosphate Bonding in High-Alumina Refractories”; Ph.D.
(1956).
9R. E. Fisher, “Hot Strength of Phosphate-Bonded Refractory Plastics,” ibid., 56 Thesis, Pennsylvania State University, University Park, 1980.
32M. Tsuhako, T. Matsuo, I. Motooka, and M. Kobayashi, “Formation of
[7] 637-39 (1977).
I0K. A. Baab and J. M. Blackwood, “Volume Stability of Phosphate-Bonded Aluminum Phosphates at Low Temperatures, "Chem. Lett., 1978, No. 7, pp. 777-80.
33J. R. Lehr, A. W. Frazier, and J. P. Smith, “Precipitated Impurities in Wet-
High-Alumina Brick,” ibid., 50 [7] 607-10 (1971).
'' F. d’Yvoire, “Aluminum andTrivalent Iron Phosphates: I, "Bull. Soc. Chim. hr., Process Phosphoric Acid," J. Agric. Food Chem., 14 [1] 27-33 (1966).
34National Bureau of Standards Mono. 24, Sec. 2, 1960.
1961, pp. 1762-76; “II,” ibid., pp. 2277-82; “III,” ibid., pp. 2283-90; “IV,”
35O. W. Florke, “Crystallization and Polymorphs of A1P0
ibid., 1962, pp. 1224-36; “V,” ibid., pp. 1237-43; “VI,” ibid., pp. 1243-46. 4 and AlP04-Si02 Mixed
12P. H. Hsu, “Interaction Between Aluminum and Phosphate in Aqueous Solu­ Crystals,” Z. Kristallogr., 125, 134-46 (1967).
36D. Lewis and M. W. Lindley, “Enhanced Activity and the Characterization of
tion, ” Advan. Chem. Ser., 1968, No. 73, 115-27.
,3M. Tsuhako, K. Hasegawa, T. Matsuo, I. Motooka, andM. Kobayashi, “Studies Ball-Milled Alumina," J. Am. Ceram. Soc., 49 [1] 49-50 (1966). H

738 Ceramic Bulletin

View publication stats