C D A U.S. Environmental Protection Agency Industrial Environmental Research	EPA-600/7-77-109
Cil	Office of Research and Development Laboratory	O # 1* AfrTT
Research Triangle Park, North Carolina 27711 o6pt©mD6r iSf /
PRECIPITATION CHEMISTRY OF
MAGNESIUM SULFITE HYDRATES
IN MAGNESIUM OXIDE SCRUBBING
Interagency
Energy-Environment
Research and Development
Program Report

-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S.
Environmental Protection Agency, have been grouped into seven series.
These seven broad categories were established to facilitate further
development and application of environmental technology. Elimination
of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The seven series
are:
1.	Environmental Health Effects Research
2.	Environmental Protection Technology
3.	Ecological Research
4.	Environmental Monitoring
5.	Socioeconomic Environmental Studies
6.	Scientific and Technical Assessment Reports (STAR)
7.	Interagency Energy-Environment Research and Development
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from
the effort funded under the 17-ageiicy Federal Energy/Environment
Research and Development Program. These studies relate to EPA's
mission to protect the public health and welfare from adverse effects
of pollutants associated with energy systems. The goal of the Program
is to assure the rapid development of domestic energy supplies in an
environmentally—compatible manner by providing the necessary
environmental data and control technology. Investigations include
analyses of the transport of energy-related pollutants and their health
and ecological effects; assessments of, and development of, control
technologies for energy systems; and integrated assessments of a wide
range of energy-related environmental issues.
REVIEW NOTICE
This report has been reviewed by the participating Federal
Agencies, and approved for publication. Approval does not
signify that the contents necessarily reflect the views and
policies of the Government, nor does mention of trade names
or commercial products constitute endorsement or recommen-
dation for use.
This document is available to the public through the National Technical
Information Service, Springfield, Virginia 22161.

-------
EPA-600/7-77-109
September 1977
PRECIPITATION CHEMISTRY
MAGNESIUM SULFITE HYDRATES
MAGNESIUM OXIDE SCRUBBING
by
Philip S. Lowell, Frank B. Meserole, and Terry B. Parsons
Radian Corporation
8500 Shoal Creek Boulevard
Austin, Texas 78766
Contract No. 68-02-1319
Task Nos. 36 and 54
Program Element No. EHE528
EPA Task Officer: Charles J. Chatlynne
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, North Carolina 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460

-------
ABSTRACT
Laboratory studies were done to define the precipitation
chemistry of magnesium sulfite hydrates. The results are
applicable to the design of magnesium-based scrubbing processes
for S02 removal from combustion flue gas, In magnesium-based
scrubbing processes, magnesium sulfite precipitates as either
the trihydrate or the hexahydrate. Equipment design and con-
ditions depend on which hydrate is formed. Theoretical pre-
dictions that were verified experimentally indicated that MgS03
trihydrate is the thermodynamically stable hydrate formed at
scrubbing process conditions, MgS03 hexahydrate is formed as
a metastable solid due to kinetic phenomena. Nucleation and
crystal growth rates are much faster for hexahydrate than for
trihydrate.
The time scales observed in kinetic experiments at scrub-
bing process conditions are: hexahydrate precipitation (tens
of minutes), hexahydrate dissolution and trihydrate precipita-
tion (hundreds of minutes), and attainment of trihydrate
equilibrium (thousands of minutes). Nucleation plays a dominant
role in the formation of trihydrate solids. These results indi-
cate that magnesium-based scrubbing processes can be designed
to precipitate a majority of either hydrate form. Important
design variables include scrubbing liquor composition and
temperature, seed crystal composition, slurry volume, equipment
residence times, and energy inputs to the slurry that influence
nucleation.
ii

-------
CONTENTS
Abstract	11
List of Figures	iv
List of Tables		
1.	Summary	 1
2.	Introduction	3
Background	3
Goals of the Study	4
3.	Conclusions	5
4.	Technical Approach and Results of the Study	8
Technical Approach 	 8
Results	9
Bibliography	61
Appendices
Technical Note 200-045-36-01, "Literature Survey of
Information Available on Magnesium Sulfite Hydrate
Formation Mechanisms" 	 63
Technical Note 200-045-36-02, "Theoretical Predictions
of Transition Temperature Lowering for the
MgS03*3H20 - MgSOj•6H2O System"	109
Technical Note 200-045-36-03, "Experimental Veri-
fication of the MgSO3 Trihydrate-Hexahydrate
Transitions at 37.5°C in MgSOi, Solutions"	123
Technical Note 200-045-36-04, "Experimental
Results for the Equilibrium Studies on MgSO3
Hydrates" 		139
Technical Note 200-045-36-05, "Experimental Results
for Precipitation Kinetics Studies on MgS03
Hydrates"	172
Technical Note 200-045-54-01, "Transition of
MgSOj Hydrates in 3M MgCla Solution"	204
Technical Note 200-045-54-02, "MgSOa Hexahydrate
and Trihydrate Preparation, Handling and
Characterization"	209
Technical Note 200-045-54-03a, "Precipitation
Kinetics of MgS03 Hydrates in Scrubber-Like
Media"	232
Technical Note 200-045-54-04, "Homogeneous
Nucleation of MgSO3 Hydrates in Scrubber-Like
Media"	309
Technical Note 200-045-54-05, "Secondary
Nucleation of MgSO3 Hydrates in Scrubber-Like
Media"	336
Technical Note 200-045-54-06, "Additive Experiments
on Trihydrate Kinetics"		 357
iii

-------
FIGURES
Number	Page
4-1 Weight Loss vs. Percent Composition for MgS03»6H20
in a Self-Generated Atmosphere	14
4-2 Solubility vs. Temperature for MgS03 Hydrates	20
4-3 Solubility Product Constant of MgS03«3H20 	 25
4-4 Intersection of Solubility Product Constants of
MgS03 Tri- and Hexahydrates Indicating the Phase
Transition Temperature in Solutions of Different
Compositions 	 26
4-5 Activity Products for MgS03 Hydrates in Solutions
in which Homogeneous Nucleation Occurred
(20 wt 7o MgSCU , pH 6)	34
4-6 Change in Sulfite Concentration in Slurry Liquor -
due to Nucleation and Crystal Growth of
MgSO 3 *6H20 	37
4-7 Activity Products for MgS03 Hydrates in Solutions
in which Primary and Secondary Nucleation
Occurred (20 wt % MgS0H and pH 6)	41
4-8 MgS03*6H20 Precipitation Rate vs Relative Saturation
at Various Temperatures	45
4-9 MgS03.3H20 Precipitation Rate vs Relative
Saturation at 55°C 	46
4-10 MgS03.3H20 Precipitation Rate vs Relative
Saturation at 70 °C 	47
4-11 MgS03*3H20 Precipitation Rate vs Relative
Saturation at 85 °C 	48
4-12 Results of MgSO, Hydrate Precipitation Rate
Experiment 3-2 Employing Trihydrate Seed
Crystals		53
4-13 Results of MgSOs Hydrate Precipitation Rate
Experiment 3-4 Employing Trihydrate Seed Crystals. . 54
4-14 Results of MgSOa Hydrate Precipitation Rate
Experiment 3-3 Employing Hexahydrate Seed Crystals . 55
iv

-------
FIGURES (Cont'd)
Number	Page
4-15 Results of MgS03 Hydrate Precipitation Rate
Experiment 3-5 Employing Hexahydrate Seed
Crystals	56
4-16 Results of MgS0 3 Hydrate Precipitation Rate
Experiment 3-6 Employing a Mixture of Tri-
and Hexahydrate Seed Crystals	58
4-17 Results of MgS0 3 Hydrate Precipitation Rate
Experiment 3-7 Employing a Mixture of Tri-
and Hexahydrate Seed Crystals	59
TABLES
Number	Page
4-1 Results of Weight-Loss Experiments for MgS03 Hydrates 15
4-2 Results of Measurements of Ion-Pair Dissociation
Constants using Two Methods	18
4-3 Measured and Published Solubility Data for
Magnesium Sulfate Hydrates 	 19
4-4 MgSC>3 Solubility Product Constants (Ksp) Calculated
from Activities of Ions in Solubility Experiments. . 22
4-5 Equilibrium Composition of MgSO* Solids in 2.3M
MgSCU Solution at 37°C 	28
4-6 Temperature and Composition of Solutions which
Produced Homogeneous Nucleation of MgSOs Hydrates. . 33
4-7 Results of Secondary Nucleation Experiments using
MgSOs*6H20 Seed Crystals 	 39
4-8 Activities and Activity Products Calculated using
Solution Data from Tests of Secondary Nucleation
of MgS03*6H20	40
4-9 Summary of Experimental Conditions in Tests of
Magnesium Sulfite Hydrate Precipitation from
Scrubber-Like Media 	 50
v

-------
ACKNOWLEDGEMENTS
The authors wish to express their gratitude to the
following people for their contributions: Dr. C. J. Chatlynne,
EPA Project Officer, for his understanding of the value of
this work; Dr. R. E. Pyle for directing the experimental portion
of the project; Messrs. R. E. Sawyer, J. L. Skloss and K. R.
Williams for performing the majority of the analytical chemistry;
and our secretaries, Mrs. Jacquie Alberstadt, Karin Weidemann,
and Dianne Sethness for typing and assembling this manuscript.
xri

-------
SECTION 1
SUMMARY
This report describes the results of work done by Radian
Corporation for EPA under Contract 68-02-1319, Tasks 36 and 54.
The work was an investigation of the physical chemistry of the
precipitation of magnesium sulfite hydrates. The study was
done to explain phenomena observed during the operation of mag-
nesium based S02 scrubbing processes. In these processes S02
is removed from flue gas by absorption in magnesium oxide/
sulfite slurries. Magnesium sulfite is precipitated as a
hydrate. Thermal decomposition of the hydrate produces mag-
nesium oxide for recycle. Operating experience indicates that
both magnesium sulfite trihydrate and hexahydrate precipitate
in magnesium-based S02 scrubbing processes.
From theoretical predictions , which were verified experi-
mentally, this study shows that the trihydrate is the
thermodynamically stable form at scrubbing process conditions.
Precipitation of the hexahydrate occurs as the result of kine-
tic phenomena. Precipitation may take place by two mechanisms:
nucleation and growth on existing crystals (crystallization).
Both nucleation and crystallization were studied. The nuclea-
tion and crystal growth rates of the hexahydrate are much
faster than the nucleation and crystal growth rates of the tri-
hydrate at conditions encountered in the magnesium oxide wet
scrubbing process.
The results of this study provide a useful design basis
for magnesium based S02 scrubbing processes. Information is
provided which defines operating conditions at which precipi-
tation of either hydrate would be expected to occur. Guide-
lines are provided which indicate how equipment design and
1

-------
selection might influence the composition of the solid hydrate
product. By controlling design variables such as slurry den-
sity and volume, seed crystal size and composition, solution
composition, and temperature, a process can be designed to pro-
duce more or less of either hydrate as the solid product.
2

-------
SECTION 2
INTRODUCTION
This report describes the results of a study of the physi-
cal chemistry of the precipitation of magnesium sulfite hy-
drates. The work was done to explain phenomena observed during
operation of magnesium oxide scrubbing processes for removing
SO2 from combustion flue gases. This explanation improves the
design basis for the S02 scrubbing system. The work was con-
ducted by Radian Corporation from August 1975 to October 1976
under EPA Contract No. 68-02-1319, Tasks 36 and 54.
BACKGROUND
The technical feasibility of the magnesium oxide wet
scrubbing process for flue gas desulfurization has been demon-
strated. The process employs a slurry of MgS03 and MgO to
scrub SO2 from flue gas. Sulfur is removed from the scrubbing
slurry by precipitation of hydrated MgS03. The hydrated MgS03
is dried and calcined to produce S02 and regenerated MgO for
recycle to the scrubbing system. Two hydrates of MgS03 are
involved, the trihydrate and the hexahydrate. Formation of
both hydrates has been observed during operation of full scale
magnesium oxide S02 scrubbing processes.
It is important for a number of reasons to be able to pre-
dict and thus control which hydrate is precipitated from the
scrubbing liquor. First, the physical properties of the two
hydrates differ. The trihydrate crystals are smaller than the
more granular hexahydrate crystals. Design of solids separa-
tion and handling equipment depends on these physical proper-
ties . Second, the drying temperatures and heat requirements
depend on how much water has to be removed to form anhydrous
3

-------
magnesium sulfite. Finally, the properties of the anhydrous
MgS03 depend on whether it was formed from the tri- or the
hexahydrate. Again, solids handling equipment design is
affected.
GOALS OF THE STUDY
The purpose of this study was to define the physical chem-
istry of MgS03 hydrate precipitation to explain why specific
hydrates are formed at various conditions. Thermodynamic
properties were used to identify the stable hydrate at equi-
librium. Kinetic studies were done to define nucleation and
crystallization rates. A further goal of the study was to show
how these physical chemistry data apply to the design of mag-
nesium-based SO2 scrubbing systems.
4

-------
SECTION 3
CONCLUSIONS
The results of this study of the physical chemistry of
precipitation of magnesium sulfite hydrates are described in
Section 4. The following conclusions are evident from the
results:
The theoretical description and analytical tools
developed and used in this investigation provided an
accurate set of results and a sound basis for the con-
clusions . The physical chemistry phenomena under in-
vestigation were first described theoretically and
then confirmed experimentally.
Magnesium sulfite trihydrate is the thermodynamically
stable phase at the conditions encountered in mag-
nesium based scrubbing processes. This result was
predicted theoretically and confirmed experimentally.
Operating experience for magnesium oxide scrubbing
processes indicates that formation of magnesium sul-
fite hexahydrate occurs at conditions for which the
trihydrate is the thermodynamically stable form.
Formation of the metastable hexahydrate occurs due to
kinetic phenomena which are described in this work.
The system under investigation is complex and a num-
ber of phenomena can occur simultaneously. These in-
clude trihydrate nucleation, hexahydrate nucleation,
trihydrate crystal growth, hexahydrate crystal growth,
and hexahydrate dissolution. These phenomena were
investigated in dilute solution experiments in
5

-------
which the rates of the processes were accurately de-
scribed. The dilute solutions were not representative
of actual scrubbing solutions.
Experiments were also done to study precipitation
rates in systems characteristic of magnesium oxide
scrubbing processes. In these systems the phenomena
described above occurred simultaneously. The results
obtained in scrubber-like systems were consistent with
the results for tests of the individual phenomena.
The following observations about the relative rates of
the individual processes are important for design con-
siderations. The rate of trihydrate nucleation at
magnesium oxide operating temperatures of 55°C appears
to be very slow compared to the rate of hexahydrate
nucleations. Although hexahydrate nucleation was ob-
served, evidence of trihydrate nucleation in scrubber-
like media was not obtained. In addition, the rate of
hexahydrate crystal growth is much faster than the
rate of trihydrate crystal growth.
The following sequence of events was observed in
scrubber-like systems and the time frame was charac-
terized: hexahydrate precipitation (tens of minutes),
simultaneous hexahydrate dissolution and trihydrate
precipitation (hundreds of minutes), trihydrate pre-
cipitation and attainment of equilibrium (thousands
of minutes).
The results of this study are directly applicable to
the design of magnesium oxide scrubbing systems. They
can provide a logical design basis for the process by
6

-------
defining the conditions under which either hydrate may
be formed.
By specifying the composition and particle size of
seed crystals, the solution composition (driving
force), and the reactor volume (residence time) it is
possible to design a system to produce either hydrate
as the solid product.
Design tools used in other SO2 slurry scrubbing pro-
cesses such as the particle balance concept are ap-
plicable to this system.
Since nucleation rate is an important factor, equip-
ment that causes high energy changes in solution will
have an effect on precipitation phenomena. Turbulence
effects of nozzles, pumping, agitator speed, and con-
trol valve throttling will be important.
Hold tank and reactor volumes and slurry density de-
termine the solid residence times. Solid residence
time in the system will be an important design
criteria.
Temperature and solution composition can be used to
control seed crystal composition.
7

-------
SECTION 4
TECHNICAL APPROACH AND RESULTS OF THE STUDY
This section of the report describes the technical ap-
proach and the results of the work to define the physical chem-
istry of HgS03 hydrate precipiation. First the technical ap-
proach is broadly summarized; then the results are described.
A complete description of results is included in the appendix.
Each work package is described in detail in a Technical Note.
Please refer to the Notes for details on methods of data col-
lection and analysis.
TECHNICAL APPROACH
The investigation of the physical chemistry of MgS03 hy-
drate precipitation included three parts. The first was to
develop analytical tools for use in data collection and analy-
sis. There were two kinds of analytical tools necessary. One
was methods for preparation, sample handling and qualitative
and quantitative analysis of MgS03 tri- and hexahydrate. The
other was a description of solution equilibria, a theoretical
framework from which to predict which species are stable under
equilibrium conditions. The description of solution equilibria
provides a means of calculating the activities of species in
solution. Activity data are used to correlate precipitation
rate data, since the driving force for precipitation is ex-
pressed as the difference between actual conditions and equi-
librium.
The second part of the technical approach was to use
thermodynamic data to predict the conditions at which the hexa-
hydrate or the trihydrate would be expected to precipitate. In
8

-------
solutions containing only magnesium sulfite in water, the hexa-
hydrate is the stable form at temperatures up to the transition
temperature of 41 °C. At higher temperatures the trihydrate is
the form that precipitates from pure solutions. The analysis
done in the second part of the technical approach showed that
the trihydrate is the thermodynamically stable species at solu-
tion compositions and temperatures characteristic of the scrub-
bing process. Thus, the formation of hexahydrate crystals had
to be explained on the basis of kinetic factors.
The third part of the technical approach was an experi-
mental study of the precipitation rates of the tri- and hexa-
hydrates. Nucleation and crystal growth rates were measured
for each hydrate separately in dilute solutions. Then tests
were done on more complex systems using solutions which more
closely approximated scrubber conditions (scrubber-like media).
Additional experiments were done to investigate the rates of
primary and secondary nucleation of the hydrates and the ef-
fects of additives on nucleation rates.
RESULTS
This section gives an overview of the work done and the
results obtained. The following subjects are discussed in sep-
arate subsections:
Summary of background information
Methods for hydrate preparation and solids handling
and characterization
Description of equilibria and measurement of equi-
librium constants
9

-------
Prediction and verification of the stable hydrate form
at scrubbing process conditions
Precipitation Rate Studies
Homogeneous nucleation of individual hydrates
Secondary nucleation of individual hydrates
Precipitation rates of individual hydrates in dilute
solutions
Precipitation experiments in scrubber-like solutions
Results are summarized in the following sections and are
presented in more detail in Technical Notes appended to the
report.
Summary of Background Information
Technical Note 200-045-36-01 provides a compilation of
background material pertinent to this study. It presents infor-
mation available from the literature on magnesium sulfite hy-
drate solubilities; transition temperatures; thermal decomposi-
tion; and methods for hydrate preparation, solids handling, and
chemical analysis. The Note also summarizes results from oper-
ating magnesium oxide scrubbing systems. Equipment employed,
chemical analyses performed, operating history, and observa-
tions of hydrates formed are described.
Methods for Hydrate Preparation, Solids Handling, and
Characterization
In order to conduct experimental investigations of the mag-
nesium sulfite hydrate system, reliable methods of sample
10

-------
handling and characterization are required. Reliable, accurate
methods preserve the chemical integrity of the sample. Such
methods assure that the material actually analyzed is represen-
tative of the solutions or slurries from which it was obtained.
Technical Note 200-045-54-02 describes these methods in detail.
The results are briefly summarized in the following paragraphs.
Pure magnesium sulfite hexahydrate is prepared by nuclea-
tion of the crystals from a magnesium sulfate solution at 30°C.
Concentrated sodium sulfite solution is added and MgS03'6H20
nucleation occurs at a sulfite concentration of approximately
0.5 moles/liter. After about thirty minutes, 10-100 ym sized
MgS03*6H20 crystals are produced.
Pure magnesium sulfite trihydrate is produced by nuclea-
tion and crystal growth from a magnesium sulfate solution at
65°C. The addition of sodium sulfite produces 1-5 um crystals.
Larger crystals, up to 50 ym in size, can be grown overnight.
Alternatively, the pure hydrates can be formed by adjusting the
temperature of a slurry of magnesium sulfite solids. The tri-
hydrate phase is obtained by overnight stirring of a slurry
maintained at 55°C. The solution can be either pure MgS03 or
20 wt 7o MgSOi, . The hexahydrate may be obtained by overnight
stirring at room temperature.
The pure hydrate crystals are separated from liquids by
vacuum filtration and alcohol washing. Fifty percent solutions
of ethanol or methanol are employed, since the solubility of
MgSO3 in the alcohol solution is 1/30 of that in pure water.
Several alcohol washings are required for complete removal of
magnesium sulfate, sulfite, or chloride solution from the
crystals. The solids are dried in an oven at 40-50°C. Higher
temperatures will cause dehydration. Thirty minutes is ade-
quate drying time for samples of one-gram or less. Dried
11

-------
solids can be stored without deterioration in a closed glass or
plastic container under nitrogen or in air of relative humidity
less than 907o.
Solids characterization methods include differential scan-
ning calorimetry (DSC), differential thermal analysis (DTA),
infrared spectroscopy (IR), and analysis of X-ray diffraction
patterns. DSC analyses measure the difference in rate of
energy absorption or evolution between a sample and a reference
material. The materials are subjected to a programmed rate of
heating in a closed system. A DuPont 990 Thermal Analyzing
System was used to develop the DSC method. A Perkin-Elmer
DSC-2 was used for analysis of the samples from precipitation
experiments. Both tri- and hexahydrates of MgS03 had only one
endothermic transition. The trihydrate transition began at
about 120°C with the maximum at about 170°C. The hexahydrate
transition began at about 70°C with the maximum at about 100°C.
The exact starting and maximum transition temperatures varied
slightly from sample to sample and between the DuPont and
Perkin-Elmer instruments, but the peaks are so widely separated
that they can be identified without ambiguity.
A DuPont 951 Thermogravimetric Analyzer was also used to
determine the composition of the hydrate mixture. The pro-
cedure involves heating a sample at a constant rate and re-
cording weight loss or gain as a function of temperature. At
100°C the hexahydrate loses three moles of water to form tri-
hydrate. At 180°C dehydration of the trihydrate to the an-
hydrous salt occurs. The percent weight loss of a mixture of
hydrates at these temperatures can be used to calculate the
weight percent of each hydrate and the magnesium sulfite
content.
12

-------
A method for characterization of solids in field applica-
tions was also developed. The procedure employs a large glass
vacuum dessicator containing a thermometer with a Variac-con-
trolled heating mantle. The temperature inside the dessicator
is adjusted to 100-110°C. At that temperature in a self-gen-
erated water atmosphere the hexahydrate is converted to the tri-
hydrate. Hydrate samples are weighed and placed in the dessi-
cator in loosely stoppered closed containers for four hours.
The samples are weighed again. The average weight loss for pure
hexahydrate samples is about 257o (theoretical weight loss is
25.67o). For trihydrate samples, the average loss is about one
percent (theoretical is zero). Results are shown in Figure 4-1
and Table 4-1.
Description of Equilibria and Measurement of Equilibrium
Constants
A theoretical model of equilibrium in aqueous ionic solu-
tions is used to calculate activities of the ions and ion pairs.
Activities are used in calculating equilibrium as well as in the
driving force term in rate data correlations. A computer pro-
gram for calculating activities based on a model of equilibrium
has been developed by Radian Corporation. The equilibrium
model is described in the literature (LO-R-007) and was used
for data analysis in this study.
In order to apply the equilibrium model to solutions em-
ployed in the magnesium oxide process, some equilibrium con-
stants had to be added. Thermodynamic data for reactions in-
volving magnesium and sodium ions were needed to extend the
model. The reactions and equilibrium constants of interest are
listed below.
13

-------
3 Or
20
80% MgSO 3•6H2O
207. MgSOs • 3H2O
o
o
M 15
90 100
7» MgS0j-6H20 in sample
Figure 4-1. Weight Loss vs. Percent Composition for
MgS03*6H20 in a Self-Generated Atmosphere
14

-------
TABLE 4-1. RESULTS OF WEIGHT-LOSS EXPERIMENTS FOR HgS03	HYDRATES
AT	100°C IN A SELF-GENERATED ATMOSPHERE
Sample Composition	—
(wtZ) Weight	(Grams)	Wt. Loss, %
Hexa Tri	Sample Wt.	Wt. Loss	Observed	Theoretical
100 0 0.599	0.152	25.4	25.4
100 0 0.602	0.153	25.4	25.4
0 100 0.602	0.0018	0.29	0.0
0 100 0.600	0.0005	0.08	0.0
60 40 0.600	0.093	15.4	15.2
60 40 0.603	0.092	15.3	15.2
80 20 0.602	0.120	19.8	20.3
80 20 0.535	0.105	17.5	20.3

-------
Magnesium sulfite ion pair dissociation constant
MgS03(aq) t Mg+2 + SO 32
Magnesium sulfate ion pair dissociation constant
MgS°,(aq) t Mg+2 + SOI2
Sodium sulfite ion pair dissociation constant
NaS0*(aq) * Na+ + SO32
Magnesium Sulfite trihydrate solubility product constant
MgS03-3H20(s) t Mg+2 + SO32 + 3H20(aq)
Magnesium Sulfite hexahydrate solubility product constant
MgS03-6H20(g) t Mg+2 + SO32 + 6H20(aq)
Technical Note 200-045-36-04 describes the experimental
approach and the procedures and analytical methods employed.
Two methods were used to measure ion-pair dissociation con-
stants. One was to measure the free cation response of a so-
lution with ion selective electrodes. These values were com-
pared with the free cation responses of standard solutions with
negligible ion pairing. The second method was to determine the
influence of ion-pair formation on the solubility of a sparing-
ly-soluble reference compound. The solubility of the reference
compound is increased by the addition of an electrolyte which
can form ion pairs with ions of the reference compound. Ion-
pair dissociation constants determined by the second method
were more accurate. It was found that selective cation
16

-------
response measurements were influenced by an electrode junction
potential inherent in each measurement. The magnitude of the
electrode junction potential, which is a function of solution
composition and temperature, is unknown but not negligible.
In the solubility influence experiments the ion-pair dis-
sociation constants were measured at temperatures of 30°C and
50 °C. CaS03'%H20 and CaSOi^HzO were used as the sparingly-
soluble reference compounds. The salts equilibrated with each
solid in individual tests were MgCl2, NaCl, and CaCl2. Chloride
salts were used because the chloride anion does not form ion
pairs to an appreciable extent with cations of the above salts.
Data collection and analysis procedures for both methods
are described in Technical Note 200-045-36-04. The resulting
ion-pair dissociation constants are listed in Table 4-2. There
were no reported values for these constants in the literature.
Solubility product constants for the tri- and hexahydrates
were calculated from the results of solubility studies for the
salts. Crystals of the pure tri- or hexahydrate were dissolved
in de-ionized water in a vessel contained in a constant tem-
perature bath and continuously purged with dry nitrogen. Hexa-
hydrate crystals were dissolved at temperatures below 40°C and
trihydrate crystals at higher temperatures. The solution was
analyzed for Mg+ , SO32, SO^2, and pH as the crystals dissolved
to monitor the approach to equilibrium. Solubilities were cal-
culated from the analytical results. Measured solubilities
are compared with literature values in Table 4-3 and Figure 4-2.
Solubility product constants at temperatures from 25 to
65°C were calculated from the activities of the ions in solution
as shown in equations 4-1 and 4-2.
17

-------
TABLE 4-2. RESULTS OF MEASUREMENTS OF ION-PAIR DISSOCIATION CONSTANTS
USING TWO METHODS
Ion-Pair
Temperature ( C)
K dissociation
Sparingly-Soluble
Ion-Specific Reference Compound
Electrode Method	Method
MgSO-,
(aq)
25
35
45
55
0.00173
0.00140
0.00113
0.00091
0.00136
0.00110
0.000896
0.000747
MgSO i
(aq)
25
35
45
55
0.00347
0.00287
0.00238
0.00196
0.00569
0.00533
0.00501
0.00473
NaSO •
(aq)
25
35
45
55
0.055
0.055
0.055
0.055
0.06611
0.05760
0.05063
0.04484
18

-------
TABLE 4-3. MEASURED AND PUBLISHED SOLUBILITY DATA FOR MAGNESIUM SULFATE HYDRATES
Temperature Stable	Measured Solubility	Reported Solubility
(°C)	Solid Phase	(g MgS03/100 g)	(g MgS03/100 g)
25.0	MgS03»6H20	0.668	0.625
30.0	MgS03»6H20	0.723	0.721
35.0	MgS03'6H20	0.861	0.835
45.0	MgS03 *3H20	0.906	0.904
55.0	MgS03 *3H20	0.814	0.790
65.0	MgS03 *3H20	0.713	0.710

-------
2.0
1.0
1.0
0.8
0.8
0.6
0.6
0.4
0.2
0.2
0
10
30
AO
0
20
50
60
70
80
90 100 110
(T'C)
Figure 4-2,
Solubility vs. Temperature for MgSOj Hydrates

-------
K
Sp G
(T) = a a	^6 / o,
Mg+2 SO3"2 H20 MgS03*6H20(s)
(4-1)
K (T) - a
sp 3	1
Mg+2 SO 3
a
3 ^ /fl
2 HzO MgS03*3H20(s)
(4-2)
The activities were calculated with the equilibrium model using
the analytical data from the solubility experiments. Input
data included concentrations of magnesium, sulfite, and sulfate;
temperature; and pH. The results are shown in Table 4-4.
Prediction and Verification of the Theiraiodynamically Stable
Hydrate Form at Scrubbing Process Conditions
Depending on solution composition and temperature, one of
the magnesium sulfite hydrates is stable at equilibrium and
would eventually be produced as the solid product. For a given
solution composition, the hexahydrate is the stable hydrate
form over the lower range of temperature. A transition to tri-
hydrate as the stable form occurs at a temperature called the
transition temperature. At temperatures above the transition
temperature, the trihydrate is the stable form at equilibrium.
In pure solutions containing only water and magnesium sulfite,
the transition temperature is 41°C. The hexahydrate is the
thermodynamically stable form below 41°; however, the precipi-
tation of hexahydrate crystals occurs at temperatures higher
than 41°C in some magnesium oxide scrubbing systems.
It has been suggested that at the solution compositions
characteristic of the magnesium oxide process, the transition
temperature might be higher than 41°C and the hexahydrate
might be the thermodynamically stable hydrate form. The valid-
ity of this explanation for hexahydrate formation was examined
using a two-step approach. First, thermodynamic properties
21

-------
TABLE 4-4. MgS03 SOLUBILITY PRODUCT CONSTANTS (Ksp) CALCULATED
FROM ACTIVITIES OF IONS IN SOLUBILITY EXPERIMENTS
Tempe rature
(°C)
Stable
Solid Phase
Ksp

0
MgS03-6H20

4.537
X
10" 5
10
MgSO 3* 6H2O

4.797
X
10~5
20
MgSO 3•6H 2O

5.368
x
10" 5
25
MgSO 3•6H 2O

5.720
x
10"5
30
MgS03-6H20

6.124
x
10" 5
35
MgSO 3* 6H2O

6.579
X
10" 5
40
MgSO3* 6H2O

7.120
X
10"5
45
MgS03-6H20
(metastable)
7.735
X
10"5
50
MgSO 3•6H2O
(metastable)
8.375
x
10"s
55
MgSO 3•6H2O
(metastable)
9.060
X
10" 5
60
MgS03'6H20
(metastable)
9.786
X
10-5
62.5
MgSO 3•6H2O
(metastable)
1.013
X
10-*
30	MgSO3• 3H20 (metastable)	1.074 x lO"*1
MgS0 3-3H20 (metastable)	8.662 x 10"5
40	MgS03*3H20 (stable)	7.120 x 10~!
^5	MgS03-3H20 (stable)	6.074 x 10"s
50	MgS03-3H20 (stable)	5.191 x 10"~5
55	MgS03*3H20 (stable)	4.466 x 10~s
60	MgSO3•3H2O (stable)	3.894 x 10~5
MgS03*3H20 (stable)	3.020 x 10~5
80	MgSO3•3H2O (stable)	2.457 x 10"s
90	MgSO3"3H20 (stable)	2.059 x 10*5
100	MgS03-3H20 (stable)	1.771 x 10"5
22

-------
were used to predict the stable hydrate form at equilibrium for
specific temperatures and solution compositions. Then experi-
ments were done in which solutions of appropriate composition
were brought to equilibrium at controlled temperatures. The
hydrate crystals were separated and characterized. The results
are described in Technical Notes 200-045-36-02, -36-03, and
-54-01.
Equation 4-3 describes the transition from magnesium sul-
fite hexahydrate to the trihydrate.
MgS03•6H20 t MgS03.3H20 + 3H20	(4-3)
In Technical Note 200-045-36-02 it is shown that this equi-
librium can be described in terms of the solubility product
constants, Ksp, of the hydrates as shown in equation 4-4.
„ . ^SOsOHaO *^0 _ Ksp6 (U ^
trans 	 		v J
^gSOs'OHjO	Ksp3
At a given temperature for a saturated solution the solubility
product constant is numerically equal to the value of the
activity product (ap). The activity products for the tri- and
hexahydrates are shown in equations 4-5 and 4-6.
ap3 = a a a3	(4-5)
Mg+ SO3 2 H20
ap6 s a a a6	(4-6)
Mg+2 S03"2 H20
ap3a3
H20
23

-------
Thus the hydrate transition can be expressed in terms of sat-
urated solution activity products.
Ktrans * ap6/ap, " «h20
(4-7)
Solutions in which the activity product ap 3 is greater
than Ksp3 are supersaturated with respect to MgS03 *31120 pre-
cipitation. Subsaturated solutions have an activity product
less than the solubility product constant. Figure 4-3 shows
the solubility product constant for the trihydrate as a func-
tion of temperature. The ordinate in Figure 4-3 is ap3.
The hexahydrate solubility properties can be plotted on
the same graph since
The lines representing the two solubility product constants on
the ap3 plot intersect at the hydrate transition temperature.
Figure 4-4 shows the two hydrate solubility product con-
stants in terms of ap3 in solutions of varying composition.
The figure shows that the trihydrate-hexahydrate phase transi-
tion temperature decreases with increasing dissolved salt con-
tent or decreasing activity of water. In 2M MgSOi, solutions,
which are characteristic of the magnesium oxide scrubbing
process, the transition temperature is predicted to be 37.5°C.
At 37.5°C the hexahydrate should be the stable MgS03 phase in
1.5M MgS0i» solution, and the trihydrate phase should be stable
in 2.3MMgSOi» solution.
Such predictions were verified experimentally as de-
scribed in Technical Notes 200-045-36-03 and 200-045-54-01.
aP3 " aPe/aH2o
(4-8)
24

-------
11.0
10.0
9.0
8.0
a.
m
7.0
Region of Supersaturation
with Respect to
MgSOj•3H,0
pj
6.0
o
V)
N
+
to
*
5.0
Region of Subsaturation
with Respect to
MgSO, . 3(1,0
4.0
2.0
120
110
100
80
90
70
60
50
30
20
0
10
T (°C)
Figure 4-3. Solubility Product Constant of MgS03*3H20

-------
ro
o\
x
5*
I
O
C4
33
n
at
X
N
o
in
«
/
11.0
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
^vo;'
Ksp MgSOa•3H20
10
20
30
40
50
60
70
80
90
100 110 120
(T°C)
Figure 4-4. Intersection of Solubility Product Constants of MgS03 Tri- and
Hexahydrates Indicating the Phase Transition Temperature in
Solutions of Different Compositions
02-2292-1

-------
A 1.5M MgSOi, solution at 37.5°C was saturated with MgS03 by
adding excess MgS03'6H20. The slurry, containing 25g undis-
solved solids, was maintained at 37.5°C in a stoppered flask
and stirred continuously with a magnetic stirrer for one week.
Chemical analysis of the solids confirmed that the hexahydrate
phase remained stable. Then 7g MgS03*3H20 were added. After
another week of stirring at 37.5°C the solids were again ana-
lyzed and found to be pure MgS03*6H20. All of the trihydrate
had dissolved and reprecipitated as hexahydrate.
Identical tests were also done in 2.3M MgSOi» solution at
37.5°C. The trihydrate phase was predicted to be the stable
one in these solutions. The MgSCU solution was saturated with
MgS03'6H20, the solution was filtered and llg of MgS03«3H20 were
added. Solids were sampled and analyzed after 4 days of stir-
ring. Only trihydrate was found. Ten grams of MgS03*6H20 were
added and solids were sampled again after 9, 15, and 16 days of
stirring at 37.5°C. The results are shown in Table 4-5. Chem-
ical, microscopic, DSC, and IR analyses confirmed that the
stable phase from the 2.3M MgSO^ solution was MgS03*3H20.
Additional tests were done to confirm the prediction that
MgS03*3H20 is the stable phase at 25°C in 3M MgCl2 solution.
These tests are described in Technical Note 200-045-54-01.
These results indicate that in magnesium oxide scrubber
liquors the hexa- to trihydrate transition temperature will be
lower than the pure solution value of 41°C. Therefore, magne-
sium sulfite trihydrate is the thermodynamically stable form,
and production of hexahydrate crystals as a metastable inter-
mediate must be the result of a kinetic phenomenon.
27

-------
TABLE 4-5. EQUILIBRIUM COMPOSITION OF MgS03 SOLIDS IN 2.3M MgSO„ SOLUTION AT 37°C
Elapsed Time
Days
Filtrate Analyses
(mole/Jt)	
J*£_
SO 3
Solids Analyses
(mmole/g)
Mr
++
SO,
Results of
Microscopic
Analysis
Conclusions
2.26
2.33
0.14
6.30
6.28
6.30
Trihydrate
Crystals
6.31 Trihydrate
Started with trihydrate
Trihydrate unchanged
Added hexahydrate
9	2.31
15
16
2.25
0.12
0.12
5.62
6.22
5.60 Mixture of tri- Mixture of tri- and
hexahydrate	hexahydrate
—	Mostly tri-		
with some
hexahydrate
6.25 All trihydrate
All hexahydrate
converted to trihydrate

-------
Precipitation Rate Studies
The investigation of kinetic phenomena involved in the
precipitation of HgS03 hydrates is based on a model of crystal-
lization. The model is useful because it describes precipita-
tion processes in terms of some elementary steps. Each step
has a rate that is dependent on some specific solution vari-
ables. The model helps explain what happens during hydrate
precipitation and provides a basis for data correlation. The
model defines precipitation, the formation of solids from a
solution of ions, in terms of the following processes:
Nucleation - the formation of a discrete solid par-
ticle where none existed before
Primary homogeneous nucleation - nucleation in a homo-
geneous solution with no other crystals present
Secondary homogeneous nucleation - nucleation in a
homogeneous solution containing seed crystals of pre-
cipitating solid
Heterogeneous nucleation - nucleation in a solution
containing foreign particles
Crystal growth or crystallization - deposition of solid
material from a solution on the surface of an existing
crystal.
According to the precipitation model, the following pro-
cesses can occur during the precipitation of magnesium sulfite
hydrates from magnesium oxide scrubbing liquors:
29

-------
secondary nucleation of MgS03*3H20
secondary nucleation of MgS03»6H20
crystallization of MgS03*3H20
crystallization of MgS03«6H20
dissolution of MgS03'6H20 crystals
Each of these processes occurs at a characteristic rate. The
rate of secondary nucleation depends on the solution tempera-
ture, composition, relative saturation, and energy input to the
solution. In laboratory experiments energy is imparted to so-
lutions by stirring. In industrial applications, pumps, noz-
zles and agitators are mechanical means of energy input. The
rate of crystal growth depends on the incorporation of magnesium
and sulfite ions into the surface of the solid. The rate is a
function of temperature and solution composition. Because the
surface growth rate is slow, diffusion from the bulk solution
to the solid surface is not a significant factor.
The approach to investigating the kinetics of the magne-
sium sulfite hydrate precipitation process was first to examine
the individual processes separately. These experiments pro-
vided an understanding of the individual processes. Then the
overall precipitation process for each hydrate was examined in
dilute solutions. Rate data correlation is straightforward in
dilute solutions, since the non-ideality is accurately describ-
ed. Finally, precipitation rates were studied in solutions con-
taining higher levels of dissolved salts (highly non-ideal so-
lutions) . These solutions are characteristic of magnesium
oxide scrubbing liquors so they are termed scrubber-like media.
Data analysis for precipitation rate studies in scrubber-like
media is more difficult because all the processes listed above
30

-------
can occur simultaneously. In addition, the model for calcu-
lating ionic activities does not yield accurate results in
solutions of high ionic strength since the parameters were
determined primarily at low ionic strengths.
The kinetics studies are described in detail in Technical
Notes 200-045-36-05, -54-03, -54-04, -54-05, and -54-06 con-
tained in the appendix. The results are summarized in four
sections: Primary nucleation, Secondary nucleation, Dilute
solution precipitation rates, and Precipitation rates in
scrubber-like media.
Primary Nucleation of Magnesium Sulfite Hydrates--
Precipitation occurs when the solubility product constant,
Ksp, is exceeded in a solution containing seed crystals. If
there are no nuclei present, metastable supersaturated solu-
tions exist. Supersaturation in a clear solution can increase
to some limit at which solids precipitate spontaneously. This
induced precipitation is called primary nucleation. Experi-
ments were done to determine solution compositions and tempera-
tures at which spontaneous nucleation of magnesium sulfite
solids occurs. Another goal of the experiments was to observe
which hydrate is nucleated and whether hydrate mixtures are
obtained.
The experiments are described in Technical Note 200-045-
54-04. The procedures involved the addition of concentrated
sodium sulfite solution to concentrated magnesium sulfate solu-
tion until nucleation was visually observed. Solutions were
stirred at a constant rate and the desired temperature was main-
tained by using a water bath. The composition of solutions from
which nucleation occurred was comparable to that of magnesium
31

-------
oxide scrubber liquors, 2.08 molal (20 wt %) MgS0H at pH 6. The
temperature range investigated was 28 to 80°C.
When nucleation occurred, the slurry was immediately fil-
tered. The solids were washed with 50°L ethanol and oven-dried
at 45 °C for 30 minutes. Oxidation of sulfite was negligible.
The solutions were analyzed to determine pH and the concentra-
tions of Mg+2 , SO32 , and SOlT2 . The solids were analyzed by DSC,
chemical analysis, microscopic examination, and X-ray diffrac-
tion. Detailed results of these analyses are given in Technical
Note 200-045-54-04.
The results of primary nucleation experiments are summar-
ized in Table 4-6. Table 4-6 shows that pure MgS03*6H20 is
nucleated between 28 and 60°C and pure MgS03*3H20 is nucleated
at 69 and 80°C. No hydrate mixtures were observed.
From the solution composition data given in Table 4-6, the
activities of Mg+2, SO^2 and H20 were calculated using the
Radian equilibrium model. These activities were used to calcu-
late the activity products ap3 and ap6 defined previously in
Equations 4-5 and 4-6. Values for ap3 and ap6/a|j q are shown
in Figure 4-5 as a function of temperature for the solutions in
which homogeneous nucleation occurred. The curves drawn in the
figure indicate the conditions at the onset of primary nuclea-
tion for the two hydrate phases. The lower set of curves in
the figure shows the same activity products for saturated solu-
tions of MgS03*6H20 and MgS03»3H20.
The ratio of the activity product, ap3 or ap6, to the ap-
propriate solubility product constant is defined as the rela-
tive saturation, RS3 or RS6. The relative saturation at which
nucleation occurs is a way of describing the driving force for
32

-------
TABLE 4-6. TEMPERATURE AND COMPOSITION OF SOLUTIONS WHICH PRODUCED
HOMOGENEOUS NUCLEATION OF MgS03 HYDRATES
Temperature

Concentration
	—, —			«—
of Ions (gmole/kg soln)
Species

°C±1
pH
Mg++
Na2SO 3
SO*
Nucleated
ap3xl0s
28
5.7
2,220
0.381
2.240
MgSOs-6H20*
30.5
40
6.0
2.190
0.393
2.206
MgS03-6H20**
31.6
50
6.1
2.028
0.519
2.053
MgS03'6H20**
35.9
60
6.2
2.091
0.602
2.112
MgS03»6H20**
37.1
69
6.2
2.040
0.678
2.065
MgS03-3H20
35.3
80
6.3
2.257
0.602
2.273
MgSO 3 * 3H20
28.8
Solution density was 1.260 g/ml at room temperature.
Rhombic crystal habit
Hexagonal crystal habit

-------
4 4.0
40.0
36.0
32.0
28.0
Activity Produces in Supar
satucacad Solution* Whan
Kuclaation Occurrad
MgSO 3.6H20
MgS03 *3HaO
P.
tO
2 4.0
O
oj
93
20.0
II <*>
O
w
16.0
so
SB
12.0
8.0
4.0
saturated
SOLUTION
ACTIVITY products
MgSOj•6Ha 0
MgS03•3Ha0
0.0 ¦
I	I

I	, |
20	30	40	50	60
TEMPERATURE (°C)
ro ao so
Figure 4-5. Activity Products for MgS03 Hydrates in
Solutions in which Homogeneous Nucleation
Occurred (20 wt % MgSOt,, pH 6)
34

-------
precipitation. A relative saturation of 3 to 4 was required
for primary nucleation of the hexahydrate. A relative satura-
tion of about 12 was required for the trihydrate.
Although magnesium sulfite trihydrate is the thermodynam-
ically stable phase in 2QTL MgSOn solutions at 40, 50 and 60 °C,
the hexahydrate is nucleated at these temperatures. Nucleation
of the trihydrate phase was observed only at temperatures above
65°C. These results indicate that a kinetic effect is involved
in the nucleation process. It was noted that the nucleation
transition temperature is higher than the equilibrium transi-
tion temperature. The rate of primary nucleation of the hexa-
hydrate is faster than that of the trihydrate. Further, pri-
mary nucleation occurs at lower values of relative saturation
for the hexahydrate than for the trihydrate.
Secondary Nucleation of Magnesium Sulfite Hydrates--
As described above, secondary nucleation occurs in solu-
tions containing seed crystals of the precipitating solid. Ex-
periments done to investigate conditions at which secondary nu-
cleation occurs for MgSCb hydrates are described in Technical
Notes 200-045-54-05 and 06.
The experimental approach involved the use of a continuous
liquid feed reactor with provisions for immediate mixing and
thorough agitation of the reactant slurry. Batch addition of
seed crystals was employed. The reactor was charged with 2M
MgSOi* - 1.1M Na2S03 solution and a known quantity of seed cry-
stals of either MgS03*3Hz0 or MgSOs *61120. Then Na2S03 was
added at a constant rate and the sulfite concentration and other
variables in the reactor slurry were monitored. Secondary nu-
cleation was investigated at temperatures from 30 to 60°C.
Technical Note 200-045-54-05 describes detailed procedures and
35

-------
equipment used for sampling, pH monitoring, chemical analysis,
and variation of stirring speed and temperature.
In order to understand methods of data collection and
analysis, it is useful to consider what happens in the precipi-
tation process. Immediately after the onset of nucleation,
an instantaneous process, the process of crystal growth begins.
At first the solid particles are very small. They become
larger as solid material is deposited on the surface of the
particles. The crystal growth rate is partly a function of the
particle surface area. At some point these changes are detect-
able as an increase in the weight percent solids. This increase
in solids can be detected visually or by measuring the weight of
solids in the slurry. Another indication of change is the sul-
fite concentration in the liquor. The sulfite concentration
decreases as sulfite ions leave the solution in the MgS03
precipitate. The change in weight percent solids and sulfite
concentration due solely to nuclei formation is probably imper-
ceptibly small since the nuclei are very small. But the crystal
growth which follows produces detectable changes in variables
which can be measured. The approach to investigating secondary
nucleation was to monitor the detectable changes produced due
to crystal growth and to assume that nucleation occurred just
prior to the point at which such changes were observed. Figure
4-6 illustrates the change in sulfite concentration in the
reaction liquid during secondary nucleation and crystal growth
of magnesium sulfite hexahydrate.
The onset of secondary nucleation for magnesium sulfite
hexahydrate could be observed visually when an increase in the
amount of solids in the slurry occurred. Nucleation in hexa-
hydrate experiments was also indicated by a decrease in sulfite
concentration in the reaction liquor. The onset of secondary
nucleation for the hexahydrate system was taken to be the point
36

-------
0.20
0.18
w
c
o
•H
4J
0)
M
¦U
(3
V
u
c
o
u
a>
u
•H
4-1
H
3
CO
0.16
0.14
0.12 -
0.10
0.08 .
Onset of Secondary Nucleation
Sulfite
Addition
as Na2S03
solution
Change in
sulfite con-
centration due
to crystal
growth during
precipitation
of MgS03 *6H20
0.06
10 15 20 25 30 35 40
45 50 55 60 65 70 75 80 85
Time (min.)
Figure 4-6. Change in Sulfite Concentration in Slurry Liquor -
due to Nucleation and Crystal Growth of MgS03*6H20
90 95
02-1573-I

-------
of naximum sulfite concentration. Data collected for the hexa-
hydrate system are summarized in Table 4-7. Solution composi-
tion data were used in the Radian equilibrium program to calcu-
late activities of the ions in solution at the onset of nucle-
ation. The activity product ap6, the solubility product con-
stant K	and the variable ap6/a„ were calculated as de-
sp6 r H20
scribed previously. These results are listed in Table 4-8. The
results are plotted as a function of temperature and stirring
speed in Figure 4-7.
It was not possible to detect the onset of secondary nu-
cleation in experiments employing trihydrate seed crystals. A
number of variations in experimental approach were employed in
an effort to detect a change in weight percent solids in the
system. These are described in Technical Note 200-045-54-05.
The results indicate that the increase in trihydrate solids con-
centration caused by secondary nucleation and crystal growth
was very small compared to the total solids concentration.
Further, the growth rate of crystals after nucleation occurred
was so slow that a decrease in sulfite concentration was dif-
ficult to detect.
An effect of stirrer speed on hexahydrate nucleation was
observed as indicated in Figure 4-7. Stirring imparts energy
to the system through crystal impact. The results indicate
that at lower stirring speeds (lower energy input) a higher
activity product (relative saturation) is required for the on-
set of secondary nucleation.
Batch experiments at 55°C were also done to investigate
the effect of additives on nucleation and precipitation rates
of magnesium sulfite trihydrate. These experiments are de-
scribed in Technical Note 200-045-54-06. MgS0i» solutions which
38

-------
TABLE 4-7. RESULTS OF SECONDARY NUCLEATION EXPERIMENTS USING MgS03*6H20 SEED CRYSTALS1
Solution Composition at Onset of Secondary Nucleation
Run
Stirring Speed
(RPM)
o
Temp.( C)
PH
[Hg^]
(gmole/kg)
[Na+]
(gmole/kg)
[so;]
(gmole/kg)
[so3]
(gmole/1
4-3B
1200
30.0
6.03
2.07
.390
2.10
.195
4-4B
1200
44.4
6.03
2.11
.690
2.14
.345
4-7B
1200
46.0
6.0
2.04
.678
2.07
.339
4-8B
1200
45.0
5.96
1.97
.704
2.00
.352
4-9B
1200
33.6
5.75
1.96
.554
1.99
.277
4-10B
1000
32.6
6.10
1.93
.514
1.96
.257
4-11B
1000
32.5
6.13
1.97
.526
2.00
.263
4-12B
1200
60.5
6.25
1.91
1.136
1.94
.568
4-12B
1200
60.5
6.25
1.91
1.060
1.94
.530
4-12B
1200
60.5
6.25
2.07
1.154
2.10
.577
4-12B
1200
60.5
6.25
2.07
1.076
2.10
.538
4-13B
1200
59.2
6.27
2.02
1.038
2.05
.519
4-14B
1000
58.3
6.28
1.96
1.062
1.99
.531
1 Tests were done by adding Na2S03 to a slurry of 0.1M Na2S03-2.0M MgSO<» containing lg of
MgS0t»*6H20 seed crystals.

-------
TABLE 4-8. ACTIVITIES AND ACTIVITY PRODUCTS CALCULATED USING SOLUTION
DATA FROM TESTS OF SECONDARY NUCLEATION OF MgS03-6H20
Run
a
a
a
K
ap6
Rel. Sat.6
ap6

Mg+2
S03'2
H20
sp6

aH20
4-3B
3.349-01
7.276-04
0.939
6.124-05
1.670-04
2.727
2.019-04
4-4B
3.182-01
1.012-03
0.9312
7.62-05
2.100-04
2.756
2.600-04
4-7B
2.955-01
1.322-03
0.9339
7.86-05
2.59-04
3.295
3.18-04
4-8B
2.900-01
1.021-03
0.9345
7.735-05
1.97-04
2.547
2.417-04
4-9B
3.034-01
7.898-04
0.9380
6.45-05
1.632-04
2.530
1.978-04
4-1 OB
2.935-01
1.053-03
0.9401
6.35-05
2.13-04
3.354
2.568-04
4-1 IB
3.006-01
1.087-03
0.9388
6.34-05
2.24-04
3.533
2.702-04
4-12B
2.498-01
1.754-03
0.9263
9.78-05
2.77-04
2.83
3.489-04
4-12B
2.531-01
1.612-03
0.9281
9.78-05
2.61-04
2.669
3.262-04
4-12B
2.798-01
1.657-03
0.9215
9.78-05
2.838-04
2.902
3.627-04
4-12B
2.833-01
1.521-03
0.9234
9.78-05
2.67-04
2.730
3.393-04
4-13B
2.749-01
1.544-03
0.9257
9.60-05
2.67-04
2.781
3.368-04
4-14B
2.621-01
1.670-03
0.9268
9.48-05
2.77-04
2.922
3.484-04
* All analyses are reported in miTh'moles/liter.

-------
44.0
O MIMAAV NUCLIAT10N IxmltMINTAL OAT*
n SICONOAAY NUCLlATtON f*M»M«NTAL OATA
JtCONOAHY NUCLIAT10N IXPffllMINTAI. OATA
" (m,»o3'|HjO) acducio stimminq inn
40.0
o
CM
24.0
12.0
a.o
PURE SOLUTION
ACTIVITY PRODUCTS
4.0
0.0
SO
30
40
20
AO
70
80
00
TBMWMATUM <*C>
Figure 4-7. Activity Products for MgSOa Hydrates in
Solutions in which Primary and Secondary
Nucleation Occurred (20 wt % MgS04 and pH 6)
41

-------
wete supersaturated in Na2S03 were stirred at a temperature of
55°C for 48 hours. Salts with metal cations of high ionic
charge were added in an attempt to influence the nucleation or
precipitation rate of MgS03-3H20. Primary nucleation was
studied in some systems and in other systems trihydrate seed
crystals were added and secondary nucleation/crystal growth was
studied. The change in solids concentration was determined
after 48 hours for solutions containing additives. This was
compared to the change in control solutions with no additives.
The results were that metal cation additives enhanced the
primary nucleation rate, but no increase in precipitation rate
was noted in solutions which contained seed crystals.
Precipitation Rates of Magnesium Sulfite Hydrates Measured
Separately in Dilute Solutions--
Precipitation rate studies were first conducted separately
for each hydrate in dilute solutions. Solution theory for
these dilute solutions is adequate to provide accurate methods
of calculating activities. Therefore, data correlation is fair-
ly straightforward and it is possible to conduct accurate, mean-
ingful rate measurements for precipitation of the individual
hydrates. The conditions in these experiments, however, are not
representative of those in actual scrubbing liquors.
The experimental procedure and data analysis methods are
described in Technical Note 200-045-36-05. The rate of solids
production and the rate of change of sulfite and magnesium con-
centrations in a batch-liquid, batch-solid stirred reactor were
measured at 20, 30, and 40°C for the hexahydrate and 55, 70,
and 85°C for the trihydrate. The procedure involved addition
of seed crystals to a 0.2M mixture of MgCl2 and Na2S03. The
concentrations of Mg+2, SOa2, and total sulfur; pH; and weight
percent solids were measured periodically thereafter.
42

-------
The precipitation rate was calculated by plotting the
amount of solid reaction product versus time and determining
the slope of the curve at various times throughout the experi-
ment. The amount of solids was determined in two ways. One
was the measurement of weight percent solids in intermittent
slurry samples. The other was from a liquid phase material
balance calculation. The decrease in sulfite ion concentration
was used to determine the number of moles of magnesium sulfite
leaving the solution through formation of solid hydrates.
Rate data were correlated based on the form shown in
equation 4-9.
R = kM(j>	(4-9)
where
R = precipitation rate, gmoles/min/gram solids
k = rate constant dependent on liquor temperature, composi-
tion, and transport parameters, gmoles/min/cm2 solid
surface
4> = driving force related to the difference between actual
and equilibrium quantities of reacting species
II = term dependent on the amount of solid phase present,
cm2/g
To attempt to account for the effect of solid surface area,
seed material was prepared by sieving the seed crystals through
a 400 mesh screen. The seed material had an average particle
diameter less than or equal to 37 ym. The measured
43

-------
precipitation rates were normalized by dividing by the mass of
crystals at the time of the measurement.
The driving force term is expressed as the difference be-
tween actual and equilibrium quantities of reacting species as
shown by the general form in equation 4-10.
ap =» activity product for the precipitating solid, ap3 or ap6
as previously defined
Kgp = solubility product constant, Ksp3 or KSp6 as previously
defined and
n = a parameter which, for this study, was set equal to the
number of cations plus anions, i.e., 2.
The driving force term can also be expressed in terms of the
relative saturation, RS, as shown in equation 4-11.
Figures 4-8 through 4-11 show measured, normalized reaction
rates in mmoles/gram/min versus relative saturation for
the temperatures studied. The figures verify that precipitation
rates for the hexahydrate are greater than those for the tri-
hydrate. The hexahydrate rate at 30°C is nearly three orders of
magnitude greater than trihydrate precipitation rate at 55°C.
~ * [<*p>1/n-Ksp1/n]n
(4-10)
where
(4-11)
44

-------
.0
.9
.8
7
6
5
4
3
2 "
1 ¦
0-
1
(30a)
(40°)
(20°)
A
4-
4-
0	1.50	2.00
Relative Saturation
Figure 4-8. MgSO$*6H20 Precipitation Rate vs
Relative Saturation at Various Temperatures
2.50
45

-------
0.0030
0.0025 •
0.0020
0.0015
0.0010
0.0005
0.0
1.00
1.50
2.00
Relative Saturation
2.50
Figure 4-9. MgS03*3H20 Precipitation
Rate vs Relative Saturation at 55°C
46

-------
o. 12 T
o.u -¦
0.10
0-09
0.03
0.07 r
0.06
0.05 ..
0.04 ..
0.03
0.02
0.01
0.0
1.00
1.50
2.00
2-50
Relative Saturation
Figure 4-10. MgS03*3H20 Precipitation Rate vs Relative
Saturation at 70°C
47

-------
0.12
0.11
0.10
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02 ,,
0.01 .
0.0
1.50
2.00
2.50
Relative Saturation
Figure 4-11. MgS03*3H20 Precipitation Rate vs
Relative Saturation at 85°C

-------
Precipitation Rate Studies of Magnesium Sulfite Hydrates in
Scrubber-Like Media—
The rate studies described in the preceding sections were
done under conditions designed to allow the investigation of
individual phenomena. The tests described in this section were
done under conditions characteristic of those in magnesium ox-
ide scrubbing processes in which these phenomena can occur
simultaneously.
The equipment and procedures described previously for rate
studies in dilute solutions were also employed in these experi-
ments. Batch addition of MgSOi, and NazSCh feedstock solutions
and MgS03 seed crystals was employed. All tests were conducted
at 55 °C in approximately 20 wt70 MgSCU solutions at pH 6. The
initial sulfite concentration and the composition of the seed
crystals were varied in each experiment. Tests were conducted
for periods of 1,100 to 11,000 minutes with periodic sampling
of the reactant slurry. Analyses performed on the slurry
liquor and solids are summarized in Technical Note 200-045-54-
03a. Table 4-9 summarizes the experimental conditions for each
test.
The data analysis methods employed are essentially the same
as those described for the rate studies in dilute solutions.
The results of sulfite analysis of the slurry liquor were used
to calculate the number of moles of MgS03 solids precipitating
from solution. Solid samples were characterized using DSC
analysis and microscopic examination.
The rate data correlation form shown in equation 4-12 is
applicable to the results of these precipitation rate studies.
Rate « ^Kinetic Factorj ^Surface Area| |Dporce8}(^-12)
49

-------
TABLE 4-9. SUMMARY OF EXPERIMENTAL CONDITIONS IN TESTS OF
MAGNESIUM SULFITE HYDRATE PRECIPITATION
FROM SCRUBBER-LIKE MEDIA1
Initial Sulfite	Composition of 1 gram
Test No. Concentration (gmole/J,)	of Seed Crystals
3-2	0.316	MgSO 3•3H20
3-3	0.475	MgSO3 *6H20
3-4	0.316	MgSOa *3H20
3-5	0.367	MgSO3 *6H20
3-6	0.374	MgS03-3H20 (0.5g) and
MgS03*6H20 (0.5g)
3-7	0.358	MgS03-3H20 (0.5g) and
MgS03*6H20 (0.5g)
Tests were done at 55°C and pH 6 in solutions containing
approximately 20 wt MgSOi,. Initial magnesium con-
centration varied from 2.01 to 2.04 gmole/S,.
50

-------
The rate equation shown in equation A-13	is derived in Technical
Note 200-045-54-03a.
_ l^f	(4-13)
dt j d(r)	^
In equation 4-13, ip is the amount of solid product at time, t,
divided by the initial amount of solids; k is the kinetic factor
which also takes into account initial crystal surface area, and
f^ is the driving force which is described in terms of the
activities of reactant species.
1/3
Precipitation rate data were plotted in the form of \p
versus time. A logarithmic abscissa was employed due to the
time scale of the reactions. The time variable used was 0 which
is equal to the logarithm of the sample time plus 10 minutes.
The ten minute period was added to the sample time so that the
first data point would not be zero.
Individual ion activities were calculated using the Radian
equilibrium model in order to describe the driving force factor.
The activities were used to calculate relative saturations.
These calculations indicated that the system approached a steady
state relative saturation of 2.5. Theoretically the steady
state relative saturation at equilibrium should be equal to 1.0.
In order to resolve this discrepancy, equilibrium experiments
were done at ionic strengths of about 4 which are characteristic
of this system. These equilibrium experiments confirmed the
order of magnitude discrepancy in calculated relative satura-
tions. These results indicate that the equilibrium model does
not provide an accurate description of equilibrium for the
system. While data correlation could be accomplished by going
through the mechanics of calculating driving force terms, the
results would be inaccurate and possibly misleading.
51

-------
The results of the precipitation rate studies in scrubber-
like media are summarized in the following paragraphs. Six ex-
periments are described. Two were done using trihydrate seed
crystals, two using hexahydrate seed crystals, and two using a
mixture of the two hydrate crystals.
Figures 4-12 and 4-13 show the results of tests 3-2 and
3-4 done with 1 gram of trihydrate seed crystals both at an
initial sulfite concentration of 0.316 gmole/£. The top parts
of the graphs indicate the fraction of trihydrate solids in the
product. Trihydrate was the only phase that precipitated in
these experiments, so the fraction of trihydrate is constant at
1.0. The rate of production of trihydrate solids was fairly
constant in both tests, although one rate was a little faster
than the other. Since the solution compositions were the same,
the driving forces would be expected to be equal in the two ex-
periments . The difference in rates is probably due to crystal
surface area effects. It appears that equilibrium was reached
1/3
in test 3-4 as indicated by the constant value of	at the
end of the test.
Figures 4-14 and 4-15 show the results of tests done using
hexahydrate seed crystals. The initial sulfite concentration
was 0.475 g/mole/Jt in test 3-3, (Figure 4-14) and 0.367 gmole/£
in test 3-5 (Figure 4-15). Hexahydrate nucleation was visually
observed in the first 20 minutes of test 3-3. Nucleation is in-
dicated in Figure 4-14 by the initial high rate of solids pro-
duction. The results of DSC solids analysis indicate that tri-
hydrate was not formed during the initial 200 minutes of this
test.
Test 3-5 (Figure 4-15) was done with hexahydrate seed
crystals at a lower initial sulfite concentration. Hexahydrate
nucleation was not apparent. Solids analyses indicate the
52

-------
Fraction
Trihydrate
5.0
•o
•H
4.0
^20 wt % MgSO^ solution
T = 55°C
Initial conditions:
3.0
CO
2.0
SO3 = 0.316 gmole/i
Mg^ = 2.02 greole/&
1.0
10
100
1000
10,000
0 = 10 + Sample Time (minutes)
Figure 4-12. Results of MgS03 Hydrate Precipitation Rate
Experiment 3-2 Employing Trihydrate Seed Crystals

-------
Fraction
Trihydrate
CO
5.0
•a
•r)
4.0
^20 wt % MgSOn solution
T = 55°C
Initial Conditions:
3.0
en
2.0
SO3 = 0.316 gmole/&
ig"1"1" = 2.02 gmole/2.
1.0
10
100
1000
10,000
6 = 10 + Sample Time (minutes)
Figure 4-13. Results of MgS03 Hydrate Precipitation Rate Experiment 3-4
Employing Trihydrate Seed Crystals

-------
U)
Ui
Fraction
Trihydrate
m
00
T3
o
at
•O
<0
¦H
u
•H
e
CO
5.0
4.0
3.0
2.0
1.0<
10
-*
-I{—f

100
-4-
1000
0 = 10 + Sample Time (minutes)
^20 wt % MgSOi, solution
T = 55°C
Initial Conditions:
pH = 6.41
SO3 = 0.475 gmole/X.
= 2.04 gmole/£
Figure 4-14. Results of MgS03 Hydrate Precipitation Rate Experiment 3-3
Employing Hexahydrate Seed Crystals

-------
Fraction
Txihydrate
2 5.0
4.0

^20 wt % MgSOi, solution
T = 55°C
Initial Conditions:
3.0
2.0
SO3 = 0.367 gmole/S-
Mg++ = 2.01 gmole/5-
1.0
2
¦»
2
6
2
6
10	100	1000
9 = 10 + Sample Time (minutes)
Figure 4-15. Results of MgS03 Hydrate Precipitation Rate Experiment 3-5
Employing Hexahydrate Seed Crystals

-------
possibility of a very small amount of trihydrate in the seed
crystals. The results of the liquid and solid analyses indicate
that the following phenomena occurred during the 3,200 minute
test. Hexahydrate precipitated during the first part of the
test. This was followed by an intermediate period during which
the amount of solids remained relatively constant. During this
period the hexahydrate dissolved and the trihydrate precipitated
After all the hexahydrate dissolved, the amount of solids again
increased due to further trihydrate precipitation. The tri-
hydrate precipitation rate was dependent on the amount of tri-
hydrate crystal surface area available. During the 930 minute
period from 470 to 1,400 minutes the fraction of trihydrate in-
creased from about 0.1 to about 1. At the end of 3,200 minutes
the product was pure trihydrate.
The results of the tests 3-6 and 3-7 employing seed cry-
stals composed of 0.5 grams of each hydrate are shown in
Figures 4-16 and 4-17. The results of tests 3-6 and 3-7 can be
explained on the same basis as those of test 3-5.
The results of tests of the precipitation rates of mixed
hydrates in scrubber-like media provide an accurate qualitative
explanation of what can occur during the operation of magnesium
oxide scrubbing processes. Data are available for obtaining a
quantitative rate correlation; however, some refinement of com-
putational tools would be required in order to calculate accu-
rate driving force terms.
The results for mixed hydrate precipitation in scrubber-
like media are consistent with those for precipitation of in-
dividual hydrates in dilute solutions and tests of primary and
57

-------
Ul
oo
Fraction
Trihydrate
o
o
•o
co
TJ
CO
H
-3-
*	»
10

-------
1 •
Fraction
Trihydrate
co
5.0
i—i
4.0
^20 wt % MgSOij Solution
T = 55°C
Initial conditions:
3.0
m
2.0
S03 = 0.358 gmole/Jl
[g** = 2.03 gmole/Jl
1.0
10	100	1000	10,000
8 = 10 + Sample Time (minutes)
Figure 4-17. Results of MgS03 Hydrate Precipitation Rate Experiment 3-7
Employing a Mixture of Tri- and Hexahydrate Seed Crystals

-------
secondary nucleation in scrubber-like media. Hexahydrate nu-
cleation and precipitation rates are faster than those of the
trihydrate. The trihydrate is the thermodynamically stable
phase. Clear evidence of trihydrate nucleation is not apparent.
These results provide an adequate basis for the design of a mag-
nesium oxide system to produce either hydrate as an intermediate
solid product.
60

-------
BIBLIOGRAPHY
LO-R-007 Lowell, P.S., et al., A Theoretical Description
of the Limestone Injection Wet Scrubbing Process,
Final Report, 2 Vols., PB 193-029, PB 193-030,
Contract No. CPA-22-69-138. Radian Project 200-002.
Austin, Texas, Radian Corporation, June 1970.
61

-------
APPENDICES

-------
TECHNICAL NOTE 200-045-36-01
LITERATURE SURVEY OF
INFORMATION AVAILABLE ON
MAGNESIUM SULFITE HYDRATE
FORMATION MECHANISMS
Prepared by:
Ralph E. Sawyer
63

-------
CONTENTS
1.	Introduction		66
Background		66
Objectives		67
2.	Basis of the Literature Survey		68
3.	Magnesium Sulfite Equilibrium and Kinetic
Data for Aqueous Systems	 70
Solubility of Magnesium Sulfite 	 70
Magnesium Sulfite in Scrubbing Liquors. ... 70
4.	Mag-Ox Plant Operating Characteristics 	 98
References	107
FIGURES
Number	Page
3-1 Effect of Temperature on Magnesium Sulfite
Solubility 	 ..... 71
3-2 Mutual Solubility of Magnesium Sulfite-Bisulfite-
and Sulfate at 40°, 50°, and 60°	74
3-3 Ternary Diagram of the Mg0-S02-H20 System (Liquid
Phase)	77
3-4 System of Components Mg(HS03)2, MgS03, H2S03, and
H20	77
3-5 Effect of Magnesium Sulfate on Magnesium Sulfite
Solubility	79
3-6 Influence of Magnesium Sulfate Concentration on
the Time of Recrystallization of Magnesium Sulfite
Hexahydrate into the Trihydrate	83
3-7 Time of Recrystallization of Magnesium Sulfite
Trihydrate into the Hexahydrate			85
3-8 Effect of Temperature and Slurry Concentration on
Conversion Rate of MgS0s*6H20 to MgS03*3H20	85
64

-------
FIGURES (Cont'd)
Number	Page
3-9 DTA Thermogram of Dehydration of MgS03»3H20 in a
Self-Generated Atmosphere	88
3-10 DTA Thermogram of Dehydration of MgS03*6H20 in a
Self-Generated Atmosphere	89
3-11 DSC Thermogram of MgS03»3H20 Dehydration in a
Nitrogen Atmosphere	90
3-12 DSC Thermogram of MgS03»6H20 Dehydration in a
Nitrogen Atmosphere	91
3-13 TGA Thermogram of MgS0 3*3H20 Dehydration in a
Nitrogen Atmosphere	93
3-14 TGA Thermogram of MgS03*6H20 Dehydration in a
Nitrogen Atmosphere	95
3-15 TGA Thermogram of MgS03*3H20 Dehydration in a Self-
Generated Atmosphere 	 96
3-16	TGA Thermogram of MgS03«6H20 Dehydration in a Self-
Generated Atmosphere . 		97
4-1	Venturi Scrubber 		104
4^2 Floating Bed Absorber	104
4-3 Boston Edison Scrubber 		105
4-4 Dickinson Station Scrubber 		106
TABLES
Number	Page
3-1 Solubility of Magnesium Sulfite in Water
(Hagisawa, 1934) 	 72
3-2 Magnesium Sulfite Hydrate Transition at Room
Temperature	80
3-3	Transition of MgS03*6H20 pH Adjusted by Diluted H2SOi»* 82
4-1	Plant Operating Characteristics	99
65

-------
SECTION 1
INTRODUCTION
BACKGROUND
In an MgO scrubber system, a circulating slurry of MgO and
hydrated MgSOa is used to absorb SO2 from a flue gas stream.
The SO2 is removed from the scrubbing system as a precipitate
of hydrate magnesium sulfite. Two hydrates have been identified,
MgS0 3*3H20 and MgS03*6H20. The crystals formed in the installa-
tion which removes SO2 from the flue gases of Boston Edison's
oil fired boiler were found to be primarily trihydrate crystals.
The installation at PEPCO's coal fired boiler produced primarily
hexahydrate crystals.
The removal of SO2 from flue gas streams by an MgO scrubbing
system has proven to be technically feasible. The status of
technology at the time the above units were designed was insuf-
ficient to predict which hydrated form would be favored. An
understanding of the phenomena involved in the production of
each hydrated form is important. The trihydrate crystals
produced are small in size. In comparison, the hexahydrate
crystals are large massive particles. Therefore, the settling
and handling characteristics of these two hydrated states are
markedly different. From a process point of view, the trihydrate
may be the desirable crystalline form because it takes less
energy to drive off the waters of hydration when the crystals
are calcined to regenerate the MgO. The economics may depend,
however, on the form of available heat. It is presently easier
from an operating standpoint to produce hexahydrate crystals
because of the ease with which the hexahydrate solids can be
removed from the liquids in the scrubber.
66

-------
OBJECTIVES
The objective of this literature survey is to present
currently available information on MgS03'3H20 and MgS03*6H20
as pure solids, in aqueous solutions, and as binary mixtures
Information on thermodynamic and kinetic factors that would
influence the system is also discussed. In addition, plant
operating data concerning slurry mixtures of MgS0 3«6H20 and
MgS03-3H20 are presented.
67

-------
SECTION 2
BASIS OF THE LITERATURE SURVEY
The following sources were consulted for the present
literature evaluation:
1.	Chemical Abstracts - This source was consulted
for the period of May 1975 - January 1977.
2.	EPA Report R2-73-244, "Conceptual Design and
Cost Study of S02 Removal from Power Plant
Stack Gas."
3.	Link, W. F., Solubilities of Inorganic and Metal
Organic Compounds Volume IX, Washington, D.C.,
American Chemical Society, p. 524.
4.	Biannual Reviews on Analytical Chemistry Applica-
tion - The reviews from the years 1971 and 1973
were considered.
5.	Battelle Memorial Institute, "Fundamental Study
of Sulfur Fixation by Lime and Magnesia," June
1966, NTIS Bulletin PB-176-843.
6.	Report No. Y-8513, "Final Report for the Evalua-
tion of the Magnesium Oxide Scrubbing System at
Potomac Electric Power's Dickerson Station,"
York Research Corp., Stamford, Conn., January 1975.
68

-------
7.	Research Center Report 5153, "Magnesia Bgse Wet
Scrubbing of Pulverized Coal Generated Flue Gas
Pilot Demonstration," The Babcock and Wilcox
Company, Research and Development Division,
Alliance, Ohio, September 1970.
8.	EPA-600/2-75-057, "The Magnesia Scrubbing Process
as Applied to an Oil-Fired Power Plant," Chemical
Construction Corporation, New York, New York,
October 1975.
9.	Other miscellaneous sources.
69

-------
SECTION 3
MAGNESIUM SULFITE EQUILIBRIUM AND
KINETIC DATA FOR AQUEOUS SYSTEMS
SOLUBILITY OF MAGNESIUM SULFITE
Pinaev (PI-070) , Link (LI-068) , Markant (MA-089) and others
present solubility data for pure MgSOa in deionized water as a
function of temperature and hydrate state. A plot of the solu-
bility of MgS03 as a function of temperature, using these data
points, is presented in Figure 3-1. The complete data set of
Hagisawa is also illustrated in Table 3-1.
MAGNESIUM SULFITE IN SCRUBBING LIQUORS
The transition temperature between MgS03'6H20 and
HgS03'3H20 is quoted as being between 40-50°C (KU-083, PI-073)
in a binary system. There is some disagreement as to where the
transition temperature in scrubbing liquors occurs. It is known
that MgS03*6H20 precipitates in the sulfur dioxide scrubber
section even when the solution temperature exceeds 50*C (MC-076).
There were at least two explanations offered for this phenomenon.
1.	Since there is * solid phase of MgO-Mg(OH>2
in the scrubber, the stable phase on the
surface of an MgO particle may be MgS03-6H2 0.
2.	The stable phase at SO2 scrubber operating
temperatures may be MgS03-6H20 and the
taoperature of transition exceeds 500C.
70

-------
O Link
A Pln««v
O Markant, at al.
i
to
S
p4
ft*
1.50

3
MgSOi*3HiO (IfoUiUbli)
I
HrS0,3H,0 (Stabla)
7J
MgSOi'6H|0 (Stable)
,50
70
T«Mp*ratur«, *C
Figure 3-1. Effect of Temperature on Magnesium Sulfite
Solubility
71

-------
TABLE 3-1. SOLUBILITY OF MAGNESIUM SULFITE IN WATER (Hagisawa. 1934)
Temp. Grains MgS03/100 Grams
(•C) Saturated Solution
Solid Phase
Tenn
cc!
>. Grams MgSO3/100 Grams
Saturated Solution
Solid Phase
0 0.338
MgS03-6H20
38

1.034*
MgSO3•3H20
15 .497
MgS03-6H20
40
Transition

MgSO]•6H2O +
25 .646
MgS03-6H20

Point

MgSO3•3Hj0
35 .846
MgSO3•6H 2 0
42

0.937
MgSO3•3H2O
40 Transition 	
MgS03•6H20 +
46

.897
MgSO3•3H20
Point
MgSO}-3H20
50

.844
MgSO3-3H20
45 1.116*
MgSO3•6H20
55

.817
MgSO3•3H20
55 1.465*
MgSOj•6HjO
65

.720
MgSO,-3H20
57.5 1.688*
MgSOj-6H20
75

.664
MgSO,-3H20
62.5 1.950*
MsSOj^H^
85

.623
MgSO3•3H20


	9£

.615
MgSO3•3H2O
*Metastable

-------
A third explanation put forth by P.S. Lowell, et al.
(LO-153) agrues that the transition temperature between
MgS03»3H20 and MgS03-6H20 solids when precipitated in aqueous
solution is dependent on the solution composition. Furthermore,
the transition temperature is lowered from a maximum of 41°C as
the dissolved solids content of the solution increases. There-
fore , the precipitation of MgS03*6H20 at scrubber conditions
cannot be explained from an equilibrium standpoint but must be
a kinetic phenomenon.
Information is also available on the chemistry of magnesium
sulfite in a slurry solution with other species such as MgSOi, •
xH20, MgO, andMg(HS03)2 (PI-072). The temperature dependence
of the solubilities of magnesium sulfite and sulfate are sum-
marized in Figure 3-2.
The chemistry of the magnesia base system plays an important
role in the absorption of SO2. The overall reaction of the sys-
tem is simply MgO + S02 H20 MgS03 •6H204-/MgS03* 3H20+.
-~
This reaction can be viewed stepwise as:
Slaking Reactions:
MgO(g) + H20 - Mg(OH)2(g)
(3-1)
Mg(0H)2(g) - Mg+2 + 20H"
(3-2)
SO3 Absorption
S02(g) +	-*¦ H2S03
(3-3)
H2SO3 + OH + HSO3 + H20
(3-4)
HSOg + OH" -~ SO 32 + H20
(3-5)
73

-------
A
500
50
300
o*.
200
II
100
300
200
100
B
. 20
III
D
Figure 3-2. Mutual Solubility of Magnesium Sulfite-Bisulfite-
and Sulfate at 40°, 50°, and 60
(A)	MgSOs Content (g/liter)
(B)	Mg(HS0j)2 Content (g/llter)
(C)	MgSOs Content (g/liter)
(D)	MgSOs Content (g/liter)
SYSTEM ZONES
(I) MgS0i*-MgSO3-H2O
(II) MgS0^-Mg(HSOS)2-H20
(III) MgSOs-Mg(HSOj)2-H20
74
02>2200-t

-------
Precipitation:
SO32 + Mg+2 Hi° MgS03-6H20(s)/MgS03-3H20(s)	(3-6)
Under actual field conditions the composition of the ab-
sorbing slurry is usually controlled by the rate of MgO make-
up. Three classes of the spray slurry are possible (DO-015).
Case 1: Excess sulfate, complete neutralization.
Composition: Mg+2, HS01, SO32, MgS03•6H20/MgS03•3H20.
Case 2: Excess MgO {or Mg(0H)2}, complete neutralization.
Composition: Mg+2, SO32, MgS03.6H20/MgS03*3H20, OH", MgO
{or Mg(OH)2)•
Case 3: Incomplete neutralization (either excess bi-
sulfite or MgO).
Composition: Mg"1"2, SO32, MgS03*6H20, MgO {or Mg(0H)2},
HSO 3 •
Case 3 is probably most prevalent because a small fraction
of the MgO is slow to react even under strongly acidic condi-
tions. On the other hand, Reactions 3-4 and 3-5 are ionic and
therefore probably instantaneous.
In the calcium oxide and calcium carbonate absorption
systems, the slurry concentration, or more explicitly, the
specific surface of the limestone is a very important parameter
and is possibly a controlling factor in S02 absorption. The
magnesia-based system shows no corresponding dependence on
slurry concentration because of the relatively high solubility
of the magnesium sulfite. The ratio of the solubility of mag-
nesium sulfite to calcium sulfite is 485 mole/mole at 78°C.
.75

-------
Since the sulfite ion acts as a buffer to maintain a nearly
constant hydroxide ion concentration, the dissolution rate of
magnesium hydroxide becomes less important. The calcium based
system has no such buffering effect. Unfortunately, other than
the developmental work done by TVA (TE-030) and several pub-
lished reports by Chemico (QU-013), the most recent information
is either considered proprietary in nature, or is as yet un-
published.
Some of the first chemical and physical data applicable
to the magnesia process came from the pulp and paper industry,
which has long used sulfite pulping with subsequent SO2 re-
covery. Because of the nature of the pulping operation, much
of this information applies to magnesium sulfite/bisulfite solu-
tions and is, therefore, more related to the clear liquor process
rather than the slurry process (MC-076). However, even in the
pulp and paper industry, there are still many characteristics
of magnesia-based solutions that have not been studied. H. P.
Markant and Associates (MA-089) began an experimental program
in the 1960's to gather data on the specific gravity, electrical
conductivity, viscosity, surface tension, solubility and SO2
vapor pressures of magnesia-based solutions over a wide range
of compositions.
A summary of this work is exhibited in Figures 3-3 and
3-4. Figure 3-3 is a ternary diagram presentation of the ex-
perimental data in the form of isothermal lines of equilibrium
composition of the saturated solutions.
A more usable form of this diagram is shown in Figure 3-4,
which is plotted in terms of the combined and free SO2. The
acid portion of the magnesium bisulfite, namely H2SO3, is
7.6

-------
Figure 3-3. Ternary Diagram of the Mg0-S02-H20
System (Liquid Phase)
BOUNDARY SYSTEM
D		 	
ABC	- M S03 M (H303)2 H20
AB	- Mg SO3 + H20
AC	- Mg (HS03)2 + H20
ADC - Mg (H803)2+H2SO3 + H2O
AD	- H9S0,+H90
A0 I 2345678
PERCENT COMBINED S02
Figure 3-4. System of Components Mg(HS0a)2,
MgSOj, H2SO3, and H20
77
02-2206-1

-------
referred to as the free S02 concentration and is by convention
considered to be one-half of the molar concentration of HSO3
expressed as S02 (DO-015).
Markant (MA-065), as early as 1965, reported the existence
of MgS03*xH20, where x is less than 3, solids at temperatures
of 59 °C and 76°C in saturated solutions. However, these sub-
stances could not be isolated because they converted immediately
to a more stable state.
Because the oxidation of sulfite can occur in a scrubbing
medium, substantial amounts of MgSOi» may be present in process
solutions. Several authors have looked into the effects of
MgSOi, upon MgSO3 characteristics in solution (PI-072; MC-076) .
In general, the addition of MgSOi, up to concentrations of 200
grams per liter have been reported to increase the solubility
of MgSO3 over temperature ranges typically encountered in MgO
scrubbing liquors (see Figure 3-5) . The pH of the MgSOi* solu-
tion was not reported in these studies.
One study noted that the concentration of MgSOi* may have
an influence on whether trihydrate or hexahydrate crystals
will form in a scrubbing slurry. Slurries were prepared with
a composition similar to conditions obtained during normal
operations of MgO scrubbing systems. These bottles were then
placed in a constant temperature bath. After a predetermined
period of time the bottles were taken out and the contents
filtered immediately, washed with methanol and dried in air at
40°C. The crystals were analyzed for sulfite, sulfate, and
MgO content. The ratio of the two hydrates was determined
using a chemical method developed at Chemico's laboratory. The
results are shown in Table 3-2.
78

-------
60° C
a 20
0
50
100
ISO
200
Magnesium sulfate concentration, g/1
Figure 3-5. Effect of Magnesium Sulfate on Magnesium
Sulfite Solubility
79

-------
TABLE 3-2. MAGNESIUM SULFITE HYDRATE TRANSITION AT ROOM TEMPERATURE
Conditions for Digestion	Trlhydrate	Hexahydrate
1.	In 8% MgSOit solution for All hexahydrate. No change.
24 hours.
2.	In 15% MgSOit solution:
a.	for 20 hours	No change.	Mostly hexa with
some trihydrate..
b.	for 64 hours	No change.	40% 6H2O and
60% 3H20.
c.	for 97 hours	No change.
3.	In 20% MgSOij solution:
a.	for 20 hours	No change.	80% 6H2O and
20% 3H20.
b.	for 64 hours	No change.	40% 6H2O and
60% 3HZ0.
c.	for 72 hours			No change.
d.	for 97 hours			No change.
80

-------
In another study the pH of saturated slurry varied while
examination time was held constant. The results of that study
are shown in Table 3-3. The pH of the medium seems to have
had some effect on the rate of transition of the hydrates. It
appears that alkaline pH retards, while acid pH favors the
rate. However, the major factor in the transition of hexahy-
drate to trihydrate would appear to be temperature.
The Russian researcher Pinaev (PI-071) noted that con-
centrations above 37o of MgSOu in a MgO scrubbing liquor are
undesirable particularly for the subsequent thermal decomposi-
tion of solids withdrawn from the system. The decomposition
of MgSOi, requires a higher temperature (1000-1200°C) than that
required to decompose MgSOa^xHaO, which is associated with an
additional consumption of heat. The MgSOi* content of the sul-
fite crystals might reach values as high as 11-13% during stor-
age of the removed crystals. He further noted that with the
addition of 0.001-0.005 wt % of p-phenylenediamine (PPDA) to
the scrubbing system, MgS03 crystals separated from the system
were nearly free of MgSOi,. The crystals were also stable dur-
ing subsequent storage.
Pinaev (PI-073) also reports that the duration of transi-
tion of magnesium sulfite hexahydrate into trihydrate depends
on the solution temperature and viscosity, which are determined
mainly by the soluble magnesium sulfate. His data (Figure 3-6)
on recrystallization times in the 50~70°C range were obtained
at a solid to liquid ratio of 1:10 with 100 and 200 grams MgSCK
per liter in solution at a pH of 6.5.
Reconversion of the trihydrate into hexahydrate can be
induced by cooling the suspension from 40 to 30°C, but the
rate of this process is several times slower than the forward
process. He reports that the time for complete conversion of
81

-------
TABLE 3-3. TRANSITION OF MgS03*6H20f pH ADJUSTED BY DILUTED H2S0i»
Conditions for DigestionResults
1.	pH 7.00: 59°C (45 minutes
In water).
2.	pH 7.00: 57°C (45 minutes
in water).
3.	pH 7.10: 60-63°C (45 min-
utes in water).
4. pH 7.00: 57-58°C (45 min-
utes in water).
5.	pH 7.00: 56-57°C (45 min-
utes in water).
6.	pH 7.15: 57-58°C (45 min-
utes in water).
7.	pH 7.40: 59-61°C (45 min-
utes in water).
All trihydrate.
Hexahydrate with few percent
trihydrate.
All trihydrate.
All hexahydrate.
All hexahydrate,
All trihydrate.
10% trihydrate and 90%
hexahydrate.
82

-------
A
10
$.0
kJ>
1.0
1
to to to to foo no
t. 0
S.0
M
3.0
T
I
"to *5 to ~ti m lii
Figure 3-6. Influence of Magnesium Sulfate Concentration
on the Time of Recrystallization of Magnesium
Sulfite Hexahydrate into the Trihydrate.
(A)	Molecules of water of crystallization
in the magnesium sulfite hydrates.
(B)	Time (minutes);
Temperature (°C): (1) 50, (2) 60, (3) 70;
MgSOi, Contents (g/liter): (a) 100, (b) 200.
83

-------
the crystalline trihydrate into hexahydrate is about 8-10 hours
at 30°C and increases to 24-26 hours at 40°C (Figure 3-7).
As the hexahydrate is the stable form of crystalline mag-
nesium sulfite at the normal temperatures (below 42°C), the
trihydrate crystals cannot be stored for a long time because
they absorb atmospheric moisture and adherent moisture and
cake into a compact mass.
Since the commercialization of MgO scrubbing of SO2 has
begun, interest has been generated in the slurry conversion to
MgS03«3H20 prior to calcination. An example of the effect of
temperature and slurry concentration on conversion from hexahy-
drate to trihydrate crystals is shown in Figure 3-8. Less
energy would be required to dehydrate MgS03*3H20 than MgS03*6H20.
As early as 1940, Peisaklov and Chertkov had noted that the
transition of MgS03 *61120 slurry to MgS03*3H20 occurred at reason-
ably low temperatures (MC-076).
More recently, studies of the thermal dehydration of the
known stable hydrates of magnesium sulfite, MgS03*6H20 and
MgS03«3H20, have been conducted. However, the reported find-
ings of the groups were apparently contradictory. Okabe and
Hari (OK-015) used three different techniques to study the
dehydration: differential thermal analysis (DTA), X-ray, and
infrared (IR). From the DTA and X-ray results they concluded
that MgS03»6H20 loses three water molecules between 60 and
100°C to form MgS03*3H20. The latter, at 200°C, completely
dehydrates to yield amorphous anhydrous MgS03. But, in the
infrared investigation, they reported that the spectrum does
not change when the trihydrate goes to the anhydrous state at
200°C. It does not seem plausible that the transformation
suggested could have occurred without any change in the infrared
84

-------
A
5.0
02 « 5 a u> n n is it zoii2
-------
spectra. The band in the 3500 cm"1 region is obviously an 0-H
stretching band, therefore, if the salt is dehydrated it would
be expected that this band should disappear.
Two other investigations using DTA were carried out by
groups at the Tennessee Valley Authority (TVA). These studies
were not published, but are presented, in part, in an EPA-
sponsored critical analysis of the magnesia process prepared by
TVA. In the above cited report, Jordan's work based upon DTA
leads to the conclusion that the thermal dehydration of MgS03-
6H20 takes place in one step starting at 100°C and that MgS03*
3H20 dehydrates in one step, starting at 160°C.
In the other study, Hatfield and co-workers reported that
MgS03*6H20 loses nearly all its water when heated in a stream
of argon or air at 104 C for 16 hours. This, too, supports the
inference that the thermal dehydration of the hexahydrate occurs
in one step. It is also reported that MgS03,3H20 is partially
dehydrated when heated in air for 16 hrs at 160°C.
In the above cited EPA report, it is suggested that the
apparently contradictory results may be due to differences in
experimental conditions. It was speculated that the samples
were heated in sealed tubes in the work of Okabe and Hori, al-
though such information was not provided in that paper. With-
out presenting a critical analysis of the effects of having the
samples open or closed to the atmosphere, the TVA report con-
cludes that the results of the TVA groups, in which the samples
are open to the atmosphere, are valid and that MgS03«6H20 de-
hydrates in one step to MgS03 at 100°C as long as the vapor
pressure of water in the atmosphere in contact with the solids
is sufficiently low, that is, lower than the equilibrium water
vapor pressure of MgSOs *31120 at 100°C.
86

-------
First, the DTA results will be considered (DA-195). A
typical thermogram for the dehydration of MgS03*3H20 is shown
in Figure 3-9. Only one endothermic transition is observed
starting at 190°C. On the other hand, in the thermogram of
MgS0 3*6H20 shown in Figure 3-10, two endothermic transitions are
observed. One starts at 90°C and the other coinciding with the
endotherm observed for the trihydrate starts at 190°C. The in-
ference to be drawn from these data is that the hexahydrate
does degrade in two steps and that the two steps involve a
transition from the hexahydrate phase to the trihydrate phase.
In the DSC studies heat input rather than the temperature
is measured. The DSC thermogram of MgS03*3H20 is shown in
Figure 3-11. Only one endothermic transition was observed
starting at 100°C with a peak maximum at 160°C. Both DTA and
DSC analyses indicate that the thermal dehydration of MgSO3«3H20
is a one step process. The DSC data indicates that the dehydra-
tion of MgS03*3H20 starts at a low temperature, 100°C, whereas
the DTA data infers that 190°C is the starting temperature.
The DSC thermogram of MgS03»6H20 is shown in Figure 3-12.
It is significant to note that only one endothermic transition
was observed starting at 45°C with a peak maximum at 90°C. DSC
results suggest that the thermal dehydration of MgS03*6H20 takes
place in one step starting at 45°C, in contrast to the two-step
mechanism starting at 90°C suggested by the DTA results. Again,
this is most probably a water vapor pressure effect.
Since the phenomenon measured by DTA and DSC is the same,
it was expected that the results would be consistent. Surpris-
ingly, this was not the case. From the DSC data, it appears
that the dehydration of both hydrates starts at a lower tempera-
ture and the hexa form dehydrates in one step and not in two
steps as was inferred from the DTA results.
87

-------
SAMPLE: MgS03-3H20
ORIGIN:
«;i7F 6 ^
ATM.
MM

RPF glass beads

T
A T
PROG. MODE heat
SCALE
SHIFT
50 •<
2
RATE
10
2. START 30 °c
o IN
_Q IN.



















































190"
f
»¦»








>




-V















f


























































1






































205 °C






































































Figure 3-9. DTA Thermogram of Dehydration of MgS03*3H20 in a Self-Generated Atmosphere

-------
SAMPLE: MgS03«6H20
ORIGIN:
SIZE 8
AT o'- "P r tPrt	MM

ref- glass beads
SCALE
SHIFT

T
A
x_.
1 'i
PROG. MODE heat
50 "J
0 IN.
i
RATE
10
i.ST
ART 30 °c

	0 IN.



r












































i°C


1
)0°C










f







(














r










1


\

1


















\ 1



















\z
205°















/



>

































J
























































Figure 3-10. DTA Thermogram of Dehydration of MgS03«6H20 in a Self-Generated Atmosphere

-------
SAMPLE: MgS03»3H20
ORIGIN:
size 12 mg
ATM.
MM

PFF ernotv Dan


T
50_	"i
_0 IN.
	A T	
1 2
ppnR MnnF heat
SCALE
SHIFT

RATE
10
.START_23_°C

0
	IN.











































<


1

>0°C





















\


/
















\


/

















\

/

















\



















\




















1






































160°



















c












Figure 3^-11. DSC Thermogram of MgS0 3»3H20 Dehydration in a Nitrogem Atmosphere

-------
1
SAMPLE: MgS03*6H20
ORIGIN:
SIZE 5 mg
ATM.
MM

REF.
PROG
RATE
empty pan
SCALE
SHIFT

J	
~0 3
A
T	
2-
. MOD
10
E heat
C
1
£ ,S7
ART 30 °c
0 IN.
0
	IN.















































°C

















v	
r-	



















—Y


































































































90°C
















	

















	



















Figure 3—12. DSC Thermogram of MgS0j«6H20 Dehydration in a Nitrogen Atmosphere

-------
The apparent contradiction between the DTA and the DSC
results can be rationalized by considering the relationship
between the dehydration reaction and the sample environmental
conditions. In any dehydration reaction, water is liberated;
if the latter is continuously removed from the reaction atmo-
sphere two consequences are observed. First, the dehydration
starts at a lower temperature, and, secondly, equilibrium is
not attained.
In the DSC studies, the sample is placed in an open dish
and is heated under a sweeping stream of nitrogen. Under these
conditions the liberated water is continuously removed from the
reaction environment. In the DTA studies, however, the sample
is heated in a self-generated atmosphere in a closed capillary
tube. Under the open conditions thermal dehydration starts at
a lower temperature, 45°C versus 90°C, and only one endothermic
transition is observed for the hexahydrate, whereas in the DTA
studies two transitions were observed.
To confirm the above rationalization of the differences
between the DTA and the DSC studies, the thermal dehydration
was studied by another independent technique, TGA, in which it
was possible to heat the samples either in an open condition or
in a self-generated atmosphere. TGA thermograms under open
conditions were obtained by the conventional procedure in which
the sample is placed in an open platinum dish under a sweeping
blanket of nitrogen. TGA under self-generated atmosphere was
achieved by placing the sample in a capillary tube with a thermo-
couple inside. Then, the whole tube was placed in the platinum
dish.
The TGA of MgS03*3Hz0 under open conditions is shown in
Figure 3-13. Thermal dehydration starts at 1008C and takes
92

-------
SAMPLE: MgS03*3H20
X-AXIS
Y-AXIS

SIZE
20 mg.

TEMP. SCALE 50 °C
inch
SHIFT 0 inch
TIME SCALE (ALT.)
SCALE 4 mg.
inch
(SCALE SETTING X 2)
SUPPRESSION 50 mo
!
t

























staz
t of
dehy<
rati
an








































































































	
































































	
	


—

	





























•"i* 
-------
place in one step. The weight loss of the sample is 34,0%,
which corresponds to the loss of 3 moles of water. TGA of
MgS03«6H20 under open conditions is shown in Figure 3-14.
Thermal dehydration of the hexaform starts at 70°C and takes
place in one step. The weight loss of the sample is 51%,
which corresponds to the loss of 6 moles of water.
The thermograms for the trihydrate and the hexahydrate,
respectively, under the conditions of a self-generated atmo-
sphere are shown in Figures 3-15 and 3-16. MgSCh *31120, as shown
in Figure 3-15, loses 34.57o of its weight in one step. This is
a very good agreement with TGA under open conditions, but ther-
mal dehydration starts at 220°C, i.e., at a higher temperature,
because of the self-generated atmosphere conditions. The weight
loss for the hexahydrate under a self-generated atmosphere, as
shown in Figure 3-16, is 51%, which corresponds to the loss of
6 moles of water. It is significant to note that the thermal
dehydration of the hexa form starts at a higher temperature and
takes place in two steps, with a weight loss of 25.570 in each
step. In other words, MgS03*6H20 loses 3 moles of water in each
dehydration step.
Thus, when the TGA study of the hexaform is carried under
a sweeping nitrogen atmosphere, equilibrium is not attained due
to the continuous removal of the liberated water. As a result,
thermal dehydration starts at a lower temperature and takes
place in one step. On the other hand, if equilibrium is ap-
proached by heating the sample in a self-generated atmosphere,
the TGA results show that thermal dehydration starts at a higher
temperature and takes place in two steps. The first dehydration
step yields MgS03*3H20 and anhydrous MgS03 is formed in the
second. These TGA observations are consistent with the explana-
tion of the differences observed in the DTA and the DSC studies.
94

-------
SAMPLE: MgS03*6H20
SIZE 20 n>g.
X-AXIS
Y-AXIS

TEMP. SCALE 50
inch
SHIFT 0 inch
TIME SCALE (ALTJ	
4
SCALE •_
inch
(SCAtE SETTING X 21
SUPPRESSION 5	mg.






















































































\



















.... -y
\



















V













































































































,


Figure 3-14. TGA Thermogram of MgS03*6H20 Dehydration in a Nitrogen Atmosphere

-------
SAMPLE: MgS03-3H20
SIZE 5 m9.
X-AXIS
TFMP SCAI P 50 °C
o inch
SHIFT U inch
TIME SCALE (ALT.)
Y-AXIS
SCALE. . 2 "59
inch
(SCALE SETTING X 2)
SUPPRESSION 60

















	











t

art i
>f de
tiydr*
tion














\



















\

























343
j
weij
;ht 1
DSS









































	








































































	





























100
150
200.
450	500
t»ftr co**ktm>« 'O* no>	at (*»q«iii wi'mi ixrwecou^iJ
TEMPERATURE*. °C
Figure 3-15. TGA Thermogram of MgS03*3H2O Dehydration in a Self-Generated Atmosphere

-------
SAMPLE: MgS03*6H20
SIZE 5 mg.
X-AXIS
Y-AXIS

TEMP. SCALE 50 °C
incK
SHIFT 0 in
-------
SECTION 4
MAG-OX PLANT OPERATING CHARACTERISTICS
In this section plant operating characteristics with exist-
ing Mag-Ox scrubber systems are presented in detail. The loca-
tion of the plant and a description of both plant and scrubber
are included. In addition, the major analytical studies per-
formed on the scrubber system and the major problems encountered
in the analyses are also discussed.
98

-------
TABLE 4-1. PLANT OPERATING CHARACTERISTICS
Plant
Contractors
Plant Description
Scrubber Description
Analytical Studies
Major Problems
Special Pilot
Plant
Babcock and
Wilcox,
National Air
Pollution
Control Ad-
ministration.
One-half MW pulverised
coal burning furnace
used to generate flue
gas. A water tube sec-
tion was used to cool the
gases to a temperature
typical of a boiler
ID
\o
Venturi Scrubber - The absorbant
slurry spray nozzle enters just
ahead of the venturi throat. The
venturi promotes gas absorption by
providing a zone with a high liquid
to gas specific surface. The gases
enter a cyclone separator where the
absorbant slurry is removed froa
the gas stream. The slurry flows
by gravity to the susip located
imediately under the cyclone.
The liquid level in the suap is
controlled by a level controller
which maintains a constant level
with treated make-up water. The
slurry composition is maintained
indirectly by adjustment of the
product flow rate which in turn
controls the water make-up rate via
the level controller.
(See Figure 4-1)
In this special pilot
plant project a number of
different analytical
studies were conducted.
The gaseous sulfur diox-
ide concentrations were
determined by coloriaetric
titration using a Barton
analyzer and by Reich test
using an iodosecric
determination. Sulfur
trioxide was determined by
condensation of sulfur
trioxide vapor into con-
densers followed by analy-
sis of the condensate.
The 540^ sampling was done
bv converting NO to
© x
nitric acid and then re-
acting the acid with
phenol disulfonic acid
to produce a yellow com-
plex which was measured
coloriaetrically. The
magnesium sulfite and
magnesium bisulfite were
determined by a modified
Palnrose analysis.
The sulfate analysis
was done by titration
with Uriim chloride after
cation removal.
In a detailed analysis of
the methods used in the
program, it was determined
that although the N'O^
sampling and analysis was
an ASTM standard method,
the procedure was tedious,
time consuming, and con-
tained a number of oppor-
tunities for experimental
error which were difficult
to subject to a numerical
analysis. Error analysis
studies and a spinning
syringe check on the Barton
titrator led to the conclu-
sion that the maxinum error
varied from 16% at 10 ppa
to 1.6* at 1000 pprc. The
sulfur trioxide analysis
did have problems with the
condensation, but the max-
imum probable error was
17%.
(cone.)

-------
TABLE W (coot.)
Plant
Contractors
Plant Description
Scrubber Description
Analytical Studies
Major Problems
Special Pilot
Plant (cont.)
O
o
Boston Edison,
Mystic Station
Facility.
Cheaico,	Scrubber located at Unit
(York Research 6 at Mystic Station on a
Corporation. 156 MW oil-fired boiler.
Test Contrac- New ducting was installed
tor)	at the stack with dampers
in the breaching to di-
vert che flow gas to the
scrubber and then back to
the stack. During most
of the period when the
scrubber vas operating,
it vas treated all the flue
gas, about 40Z over de-
sign flow at full boiler
loads.
Floating Bed Absorber - The float-
ing bed absorber Includes a sump,
two contact stages, and a liquid
disengagement section. Above the
smp» the FBA consists of tvo
stages. Each stage is packed with
6-8 inches of "viffle balls." The
spray nozzles located above the
top tray directs the spray of
absorbing slurry onto the top tray.
Since the gas rises through the
slurry it coses into intimate con-
tact with the slurry. Flow to the
spray nozzle is controlled and the
slurry conposition in the simp is
controlled by the product nozzle
flow. (See Figure 4-2)
The scrubber is a single venturi
scrubbing system. The MgO slurry
enters the upper part of the ab-
sorber with the untreated flue gas.
The gas and slurry mixture pass
through the throat area into the
diverging section, then into a
central dovncoaer to exit the ves-
sel. The flow of cleaned flue gas
turns 180° upward. A bleed stream
of slurry is centrifuged to remove
the solids, and the mother liquor
is sent back to the absorber.
(Sec Figure 4-3)
The sulfate analysis was
done by titration with
bariua chloride after
cation removal.
The modified Palnrose
analysis of the magnesium
sulfite species had errors
arising from volume errors
in the syringe, titration
errors, and errors in the
chemical reagents. The
sulfate analysis suffered
from a vague end point
when done titriaetrically.
When they were performed
gravinetrically, sasple
loss was a major problem.
The analytical studies
performed the determina-
tions of MgO by acid
base titration; the de-
termination of SO a in
MgS03 iodoaetrically; the
determination of S0», in
MgSOj, nepheloretrically;
the determination of car-
bon in a Lero combustion
furnace; and several types
of noisture analysis of
the free and combined
water. Other analyses
performed included iron
by coloricetric deter-
mination; nickel deter-
aination by dinethyl gly-
oxine method; vanadium
determination spectro-
photometrlcally; chloride
A thorough analysis of
analytical problems en-
countered in the program
was oiade by York Research
personnel. It appeared
that the determination of
X MgO by acid base titra-
tion had few sources of
experimental error. In
the determination cf
in MgS0 3 by the iodine
sodium thiosulfate nethod
the greatest source of
error is undissolved saaple
not reacting with the
iodine in solution. In
the deterrjin.it ion c:' SC. in
MgS0„ by nephelor-.etry che
greatest source of error
revolves around experimen-
tal aad individual error;
(cent.)

-------
IASLZ 4-1 
-------
TA*L£ 4-1
Plant
Contractors
Plant Description
Potoaac Electric
Power Company,
Dickinson Station
Cheaico,
(York Research
Corporation.
Test Contrac-
tor)
Scrubber located at tTnic
3 at Dickinson Station
is 185 MW pulverized coal
boiler. The system was
arranged so that gases
entering the scrubber can
pass through an electro-
static precipitator or
bypass it and go directly
to the scrubber.
O
N>
Severo-Donetsk NIIOGAZ (USSR)- (Not Known)
Power Plant.	State Scien-
tific Re-
search Insti-
tute on
Industrial
and Sanitary
Purification
of Cases.
Scrubber Description
Analytical Studies
Major Problems
The scrubber is a two-stage venturi
scrubbing system which had a design
capacity of handling about 501 of
the flue gas from Unit 3. In the
first stage the flue gas is contac-
ted with recycle liquor in a wet
venturi to remove particulate.
The wet gas is passed through a
series of alst eliminators and they
enter the second stage in a fixed
throat venturi. The MgO in the
second stage recycle liquor reacts
with the SO2 and forms a MgSOj
slurry. A bleed stream is sent to
the centrifuge where the solids are
removed and the nother liquor sent
back to Stage 2 for recycling.
(See Figure 4-4)
(Xot Known)
The c'necical analyses
were similar to the study
performed by Cheaico on
magnesia based SO2 scrub-
ber systems at Boston
Edison's Mystic Station.
There were a few excep-
tions: (1) the acid con-
centrations used in the
X 5ig0 cetera!naric-c vere
increased; (2) t^.e pene-
tration analysis on che
centrifuge cake was
deleted; (3) the combined
water analysis perforced
on the 3EAUS aoisture
balance was now done at
higher temperature of
160°C.
A ausber of analytical
studies were performed.
These analyses it.ciuced:
1.	concentration of SO;
in gas:
2.	ratio of solid portion
of suspension to its
liquid phuse;
3.	sdia-liquid (vt) cf
w working suspension;
A. content of solid phase
in easion;
5.	aciditiy of working
suspension {p'r.j of
sole tier.;
6.	density of working
suspension and solu-
tiom.
Several major problems
that were encountered
include the formation of
MgSO3.6H20 in the scrub-
ber slurry after scrubber
operations of 10-12 hours
when MgS0 3.3Hz0 was the
expected stable product.
This created problems
with solids handling in
the system. In addition,
the reactivity of regen-
erated MgO decreased when
it was added to the system
The analyses outlined
appeared to be well de-
veloped and effective.
The crystallizjtion mois-
ture of MgSOj is deter-
ained by a weight loss
method on drying ot the
crystals, but some details
were lost in translation.
Further work proposed at
the power plant include:
(1)	determination of the
efficiency of additives
of oxidation inhibitors in
tests of absorbers with
replaceable spherical pack-
log on che pilot equipment
(2)	further studlei of low
(cont.)

-------
7A3LE 4-1 (coot.)
	Plant	Contractors	Plant Description
Severo-Donetsk
Power Plant
(ccmt.)
Scrubber Description
Analytical Studies
Major Probleass
7.	viscosity of working
suspension and solu-
tion.
8.	aoisture of solid
phase (MgSOa crys-
tals) : hydroscopic
and crystalline,
9.	chemical composition
of liquid phase,
10.	chemical composition
of solid phase,
11.	chemical composition
of sagaesite.
teoperature oagn.esite
scrubbing with HgS0;.6H20
precipitation,
(3) further studies of high
temperature magnesite
scrubbing with MgSOj.
3HZ0.

-------
.VbnEX ELIMUfcKMS
UCUCUAT10N
n»#
Figure 4-1. Venturi Scrubber
ODflSTEA SECTION
OUST SMCLINC
LOCATION
rATICH
VALVT
RSCIKUUMION
MA 111
Figure 4-2. Floating Bed Absorber

-------
ANNULUS	MtET
CONE
•ASM
TANGENTIAL
WASH —~r
CONE
ANNULUS
CLEAN GAS
OUTLET
TO STACK
		
MIST
_ ELIMINATOR
NORMAL LIQUOR LEVEL-
rtiMC
SUCTION
Figure 4r>3, Boston Edison Scrubber
105"

-------
TO DRY ASH
HANDLING
SYSTEM

-------
REFERENCES
DA-195 Dauerman, Leonard, Sulfur Oxide Removal from Power
Plant Stack Gas. Magnesia Scrubbing Regeneration:
Production of Sulfuric Acid, Newark, N.J. , Inst, of
Technology, Feb. 1975.
DO-015 Downs, W., and A. J. Kubasco, Magnesia Base Wet
Scrubbing of Pulverized Coal Generated Flue Gas--
Pilot Demonstration, Alliance, Ohio, Babcock and
Wilcox Co., 1970; PB 198-074.
KU-083 Kuzminykh, I. N. , and M. D. Babushkina, "Equilibrium
Between Sulfur Dioxide and Magnesium Bisulfite Solu-
tions," J. Appl. Cheia. USSR 30(3) , 495-98 (1957).
LI-068 Linke, William F., Comp., Seidell's Solubilities;
Inorganic and Metal-Organic Compounds, 4th ed.,
2 vols., Princeton, N.J., D. Van Nostrand, 1958
(Volume 1) and 1965 (Volume 2).
LO-153 Lowell, P. S., Formation of Tri- and Hexahydrates of
Magnesium Sulfite, Technical Note 500-403-03, Austin,
Texas, Radian Corporation, December 1974.
MC-076 McGlamery, G. G., Tarstrick, R, L.; Simpson, J. P.,
and Phillips, J. F., Conceptual Design and Cost Study.
Sulfur Oxide Removal from Power Plant Stack Gas.
Magnesia Scrubbing - Regeneration: Production of
Concentrated Sulfuric Acid, EPA-R2-73-244. Muscle
Shoals, Ala., TVA, 1973.
107

-------
MA.-089 Markant, H. P., N. D. Phillips, and I. S. Shah,
"Physical and Chemical Properties of Magnesia-Base
Pulping Solutions," Tappi 48(11), 648-53 (1965),
OK-015 Okabe, Taijiro, and Shoichiro Hori, "Thermal Decomposi-
tion of Magnesium Sulfite and Magnesium Thiosulfate,"
Technol. Repts. Tohoku Univ. 23/2), 85-9 (1959).
PI-070 Pinaev, V. A., "The Viscosity and Density of Magnesium
Sulfite-Bisulfite-Sulfate Solutions," J. Appl. Chem.
USSR 36(10) , 2253-55 (1963).
PI-072 Pinaev, V. A., "Mutual Solubility of Magnesium Sulfite,
Bisulfite and Sulfate," J. Appl. Chem. USSR 37(6),
1353-55 (1964).
PI-073 Pinaev, V. A., "Recrystallization of Magnesium Sulfite
Hydrates," J. Appl. Chem. USSR 43(4), 869-70, (1970).
QU-013 Quigley, Christopher P., "Operational Performance of
the Chemico Magnesium Oxide System at the Boston Edison
Co., Part II," Presented at the Flue Gas Desulfuriza-
tion Symposium, New Orleans, 14-17 May 1973.
TE-030 Tennessee Valley Authority, Removal of Sulfur Dioxide
from Stack Gases. Thermal Decomposition of Magnesium
Sulfite, Muscle Shoals, Ala., Feb. 1971, #38.
WA-065 Walden Research Corp., Improved Chemical Methods for
Sampling and Analysis of Gaseous Pollutants from the
Combustion of Fossil Fuels; Vol. 1, Sulfur Oxides,
by J. N. Driscoll and A. W. Berger, Final Report,
Contract No. CPA 22-69-95. Cambridge, Mass., 1971.
108

-------
TECHNICAL NOTE 200-045-36-02
THEORETICAL PREDICTION OF TRANSITION
TEMPERATURE LOWERING FOR THE
MgSO3.3H20 - HgSO3.6H20 SYSTEM
Prepared by:
R. E. Pyle
P. S. Lowell
109

-------
CONTENTS
1.	Introduction and Summary 		m
2.	Theoretical Predictions	112
3.	Conclusions	121
4.	Nomenclature	122
FIGURES
Number	Page
2-1 Solubility Product Constant of MgS03*3H20	113
2-2 Transition in MgSCK Solution	115
2-3 Transition in NaCl Solution	120
TABLES
Number	Page
2-1 Saturated Conditions for MgS03*6H20 in Pure Water. . 116
2-2 Influence of MgSOi, Addition	117
2-3 Influence of NaCl Addition	119
110

-------
SECTION 1
INTRODUCTION AND SUMMARY
The magnesium oxide (Mag-Ox) process for removing S02 from
flue gases has been applied to several power plants. While the
results have generally been successful, there are still several
areas of uncertainty. One is the inability to predict which
hydrate of magnesium sulfite is precipitated at a given set of
conditions.
There has been speculation that in the solutions encountered
in actual operating scrubbers the transition temperature for
MgS03.3H20 and MgSO3.6H20 is raised. Mag-Ox units treating
flue gases from oil-fired units tend to give trihydrate while
coal-fired units give hexahydrate.
In this technical note the theoretical framework for predict-
ing the transition temperature is developed and numerical results
are calculated. Theory predicts that in all practical solutions
the transition temperature is lowered from the ^ 41°C for pure
solutions. In most actual scrubber liquors a transition tempera-
ture lowering of 3 to 10 degrees may be expected. Therefore,
the effect of solution composition on the transition temperature
cannot explain the existence of MgS03.6H20 under scrubber operat-
ing conditions.
An experimental verification of the theoretical predictions
is given in Technical Note 200-045-36-03.
Ill

-------
SECTION 2
THEORETICAL PREDICTIONS
In Technical Note 500-403-03, it was shown that the transi-
tion equilibrium constant could be obtained from the solubility
product constants of the hydrates. That is:
KSPJ -	(2_la)
»+2flsnT a6/a..._„«	r\	(2-lb)
Kspe * ®Mg aS0a a£/aMgS03 .6H20
Ktrans * Kspe/Ksp3	(2-2a)
= a3
w ^gSOsOHzO^gSOa-eHzO	(2-2b)
" %	(2-2c)
It is convenient to plot solution properties versus tempera-
ture in a general fashion. A means of doing this for the trihy-
drate is to plot the activity product,
apa 5 a14g+!aso;2a^	(2-3)
At a given temperature, the saturated solution value of
ap3 is the solubility product constant, Ksp3. This value is
plotted in Figure 2-1. Solutions with values of ap3 greater
112

-------
11.0
10.0
9.0
8.0
a
a>
7.0
Region of Saturation
with Respect to
MgS03*3H20
o
CM
as
6.0
ca
X
5.0
>
Region of Subsaturation
with Respect to
MgS03-3H20
4.0
3.0
2.0
120
100
80
110
60
90
AO
50
70
20
10
30
0
T (°C)
Figure 2-1. Solubility Product Constant of MgS03*3H20
02-2294-1

-------
than Ksp3 are supersaturated with respect to the trihydrate,and
solutions with ap3 less than Ksp3 are subsaturated.
It would also be convenient to plot the solubility tendencies
of the hexahydrate on the same graph as the trihydrate species.
The property of interest for the hexahydrate is its activity pro-
duct, ap6. That is:
ap6 = a., ~^~2 a0l2 a6	(2-4a)
F6 Kg S03w '	v '
- ap3 .	(2-4b)
If we consider the ordinate on our ap3 plot as y, then we can
plot ap6 values on our ap3 graph by defining the y ordinate as:
ape = y a^, or	(2-5a)
y = ap 6/a^	(2-5b)
Different saturated solutions of hexahydrate will now be
plotted as a single line on the ap3 graph. For solutions composed
of pure MgS03, the saturated condition for hexahydrate may be
plotted on the ap3 graph from the data given in Table 2-1. This
is shown in Figure 2-2 as the solid line.
Also plotted on Figure 2-2 are values of Ksp6/a^ for MgS03
in 1.0 and 2.0 M MgS04 solutions. The data are taken from
Table 2-2.
114

-------
* /v/.0 /
/'•

Ksp MgS03'3H20
10
20 30 40 50
60 70
(T°C)
_L
_l_
-L
X
80 90 100 110 120
Figure 2-2. Transition in MgSO* Solution

-------
TABLE 2-1. SATURATED CONDITIONS FOR MgS03.6H20 IN PURE WATER
Temperature
(°C)	Ksps
25	5.720x10"5
40	7.120x10"5
45	7.735x10"5
55	9.060xl0~5
62.5	10.13x10*5
aw	y = KSp6/a^
0.9987	5.742x10"5
0.9982	7.161x10"5
0.9978	7.786x10"5
0.9971	9.139x10"5
0.9964	10.24x10"5
116

-------
TABLE 2-2. INFLUENCE OF IlgSO, ADDITION
aw	y -
Temp.		 		
(°C)	KspexlQ5	1 M 2 M 1 M	2 M
25	5.720	0.932 0.858 6.14xl0~5	6.67xl0~5
40	7.120	.932 .858 7.64x10"=	8.30xl0"5
45	7.735	.932 .858 8.30x10"5	9.02X10'5
55	9.060	.932 .858 9.72xl0"5	10.56xl0"5
62.5	10.13	.932 .858	10.87xl0"5	11.81xl0"5
Note: 1 M MgSO,, = 11 vt 1 MgSO.,
2 11 MgSO,, = 19 wt 7o MgSO4
117

-------
The intersection of these lines on the ap3 plot (Figure 2-2)
is the transition point. From this it is seen that the trihydrate-
hexahydrate phase transition point continually decreased with in-
creasing dissolved salt content or decreasing activity of water.
Sodium chloride has a more significant effect than magnesium
sulfate as may be seen from the data in Table 2-3. These data
are given graphically in Figure 2-3.
118

-------
TABLE 2-3. INFLUENCE OF NaCl ADDITION
Temp.
(°C)
25
40
45
55
62. 5
KsPfiXlQ5
5.720
7.120
7.735
9.060
10.13

1 M
2 M
0.905	0.815
.905	.815
.905	.815
.905	.815
.905	.815
y - Ksp6/a^
1 M
2 M
6.32xlO~5	7.02x10"5
7.87x10"5	8.74x10"5
8.55x10"5	9.49x10"5
10.01x10"5	11.12xl0"5
11.19x10"5	12.43x10"5
Note: 1 M NaCl
2 M NaCl
5.5 wt % NaCl
10.5 wt % NaCl
119

-------
11.0
10.0
9.0
8.0
7.0
.0
5.0
4.0
3.0
2.0
60
100 110 120
30
50
70
80 90
20
40
0
10
(T°C)
Figure 2-3. Transition in NaCl Solution
02-2297-

-------
SECTION 3
CONCLUSIONS
From this description it can be seen that theory predicts a
MgS03 hexahydrate-trihydrate phase transition temperature lower-
ing with increasing solution dissolved solids content.
This indicates that for all operating units to date
(T >37°C), the trihydrate is the stable form. Production of
hexahydrate crystals as a metastable intermediate must be a
kinetic phenomenon.
121

-------
SECTION 4
NOMENCLATURE
a	activity
K	equilibrium constant
y	activity product ordinate defined in Equation 2-5b.
Subscripts
p	product
trans	transition, hexa to tri
w	water
3	trihydrate
6	hexahydrate
122

-------
TECHNICAL NOTE 200-045-36-03
EXPERIMENTAL VERIFICATION OF THE MgS03
TRIHYDRATE-HEXAHYDRATE TRANSITIONS
AT 37.5°C IN MgSO., SOLUTIONS
Prepared by:
J. L. Skloss
123

-------
CONTENTS
1.	Introduction	125
2.	Experimental	126
1.5M MgSOi* Solution Test	126
2. 3M MgSOit Solution	127
3.	Results	128
4.	Conclusions	138
FIGURES
Number
3-1 MgS03-6H20 Crystals, 100X	129
3-2 MgS03 • 3H20 Crystals, 200X	 129
3-3 DSC Scan of the MgS03 Solids Present in the 2.3M
MgSOt* Solution	132
3-4	DSC Scan of Pure MgSO3*3H20	133
3-5	DSC Scan of Pure MgS03*6H20	134
3-6	DSC Scan of the Intermediate MgSOa Solids	135
3-7	IR Spectra of Pure MgS03*3H20 and Pure MgS03*6H20 , . .	136
3-8 IR Spectra of the MgS03 Solids Present in the 2.3M
MgSOt* Solution and Pure MgS03*3H20. 		137
TABLES
Number	Page
3-1 Transition of MgS03 Solids in 2.3M MgSO^ Solution
at 37°C	130
124

-------
SECTION 1
INTRODUCTION
It has been shown theoretically that the transition tem-
perature 41°C in pure solutions) of MgS03*3H20 and MgS03*6H20
can be lowered by increasing the ionic strength of the solution.
In Technical Note 200-045-36-02 it was predicted that a 2.0M
MgSOu solution could lower the transition temperature to 37.5°C.
Therefore, the hexahydrate phase of magnesium sulfite is pre-
dicted to be stable in 1.5M magnesium sulfate solution at 37.5°C,
whereas the trihydrate phase is predicted to be stable in 2.3M
magnesium sulfate solution at the same temperature. It was de-
sirable to verify these predictions by experiment.
125

-------
SECTION 2
EXPERIMENTAL
Magnesium sulfate solutions (1.5 and 2.3M) were prepared
and heated to 37.5°C. Excess hexa- and/or trihydrate magnesium
sulfite crystals were added to these solutions and stirred con-
tinuously for several days. Sample aliquots were filtered, and
the hydrate state of the magnesium sulfite solids was determined
by chemical analyses and microscopic examination of the solids.
In the case of the 2.3M magnesium sulfate solution test, dif-
ferential scanning calorimetry (DSC) and infrared (IR) scans of
the solids were also performed.
1.5M MgSOi* SOLUTION TEST
A 1-liter solution of 1.5M magnesium sulfate was prepared
and heated to 37.5°C. Excess MgS03*6H20 solids were added to
saturate the solution with respect to magnesium sulfite and to
leave 25g of solids undissolved. The 1-liter erlenmeyer flask
containing the slurry was stoppered and placed into a water bath
maintained at 37.5°C. The slurry was stirred continuously with
a magnetic stirrer. The hexahydrate phase remained stable dur-
ing a one-week trial period as confirmed by chemical analyses.
Then 7g of MgS03 *31120 solids were added. After one week of
stirring at 37.5°C, a 100-ml portion of this slurry was filtered
on Whatman No. 42 paper. The filtrate was saved for chemical
analyses, and the solids were washed with 507o ethyl alcohol solu-
tion preheated to 37°C. The solids were oven dried at 45°C for
30 minutes, then analyzed and found to be pure MgS03*6H20.
126

-------
2.3M MgSOi» SOLUTION
500 ml of 2.3M magnesium sulfate solution was prepared and
heated to 37.5°C. Sufficient magnesium sulfite hexahydrate
solids were added to saturate the solution. After filtration,
llg of MgS03*3H20 were added. A nitrogen purge was used each
time the erlenmeyer flask containing the slurry was opened.
The solids were allowed to equilibrate with the 2.3M mag-
nesium sulfate solution for four days with gentle stirring in a
water bath set at 37.5°C. A 50-ml sample was withdrawn and
filtered by the same technique as described in the 1.5M MgSOu
Solution Test Section. The solid sample was analyzed to be
pure trihydrate.
Then, lOg of MgS03*6H20 was added to the 2.3M magnesium
sulfate solution (which contained the MgS03*3H20 solids). The
slurry was stirred continuously for five days at 37.5°C. A
50-ml sample was withdrawn and the slurry was allowed to equili-
brate for another week under the same conditions. Finally,
another sample was taken.
Chemical analyses of the filtrate and solids and microscopic
examination of the solids were performed each time the slurry was
sampled. The solids from the final sample were also analyzed by
the DSC and the IR methods. The stable phase from this solution
was found to be MgS03'3H20.
127

-------
SECTION 3
RESULTS
The chemical analyses of the final 1.5M magnesium sulfate
solution containing the magnesium sulfite solids showed that
the filtrate contained 1.5M magnesium and 0.13M sulfite. The
solids contained 4.71 ±0.20 mmole/g of magnesium and sulfite
which indicated that the solids were pure MgS03*6H20 and that
all the MgS03*3H20 solids were converted to the hexahydrate
phase of magnesium sulfite. Microscopic examination also
showed that the solid sample was MgS03-61120. The hexahydrate
phase was observed to consist of relatively large hexagonal
crystals (see Figure 3-1).
The chemical analyses of the 2.3M magnesium sulfate solu*
tion containing the magnesium sulfite solids are shown in
Table 3-1. The magnesium and the sulfite concentrations of the
filtrates were fairly consistent. The magnesium and the sulfite
analyses of the solids showed that the trihydrate phase of mag-
nesium sulfite remained stable during the four-day trial period.
After the addition of MgS03*6H20 solids to the slurry, the
magnesium and the sulfite content of the solids dropped. Anal-
yses showed a mixture of phases was present in the slurry after
five days of continuous stirring. Within the next 12 days, an
increase in the magnesium and the sulfite content to near the
theoretical value of 6.31 mmole/g indicated that the hexahydrate
phase had converted to the trihydrate phase. Microscopic ex-
amination also showed that the final solids sample was MgS03*3H20.
The trihydrate phase was characterized as being relatively small
bipyramid crystals (see Figure 3-2) .
When the DSC scan of the final solids sample, shown in
Figure 3-3, was comapred with the DSC scans of the pure
128

-------
C% r ^
^ •
FIGURE 3-1. MgS03 6H20 CRYSTALS. 100X
%
FIGURE 3-2. MgS03 3H20 CRYSTALS. 200X
129

-------
TABLE 3-1. TRANSITION OF MgS03 SOLIDS IN 2.3M MgSO* SOLUTION AT 37°C
Total Time
Allowed,
Days
Filtrate Analyses
	,(mole/ft)			
Mg2 so3~2
Solids Analyses
	Cmmole/g)
Mg
Sulfite
Microscopic
Analysis
Conclusions
2.26
6.30
6.30
Trihydrate
Crystals
Started with trihy-
drate
2.33
0.14
6.28
6.31
Trihydrate
Trihydrate unchanged
Added hexahydrate
9	2.31
0.12
5.62
5.60
Mixture of tri-
hexahydrate
Mixture of tri- and
hexahydrate
15
Mostly tri-
with some
hexahydrate
16
2.25
0.12
6.22
6.25
All trihydrate
All hexahydrate
converted to trihydrate

-------
trihydrate and the pure hexahydrate solids (Figures 3-4 and
3-5, respectively), it was concluded that the final solids
sample was the pure trihydrate phase. The loss of water (or
the surge of energy required to maintain a steady temperature
increase) started at 160°C and obviously continued well beyond
the 80°C temperature needed for the hexahydrate phase.
Figure 3-6 shows the DSC scan of the solids from an earlier
sample five days after the addition of hexahydrate crystals,
when both phases of magnesium sulfite were present. Note the
two characteristic peaks for the hexa- and the trihydrate
phases.
IR spectra of pure magnesium sulfite trihydrate and hexa-
hydrate were also made (see Figure 3-7). Small differences in
these spectra can be noticed. The IR spectrum of the final
solids sample from the 2.3M magnesium sulfate solution is shown
in Figure 3-8 where it is compared with that of pure magnesium
sulfite trihydrate standard. These spectra matched well and it
was concluded that the solids from the 2.3M magnesium sulfate
solution were all trihydrate. Sulfate, if present, would ab-
sorb at 1,100 cm"1. Since no absorbance was observed in this
region, it was concluded that the equilibrated solids sample
was sulfate free.
131

-------
PAtr NO. fMOft
T-AXIS
DTA-DSC
TMA
BUN NO	DATE
SCALE. "C/tn J O
t mcal/sec )/in
WEIGHT, mg 4. 3 3
SCALE. C/». 20
PROG. RATE. !C/min_
HEAT	COOt	ISO
SHIFT, in	±S
OPERATOR
ATM Pin.
SCALE, ma/in
SCALE, mils/in
MODE
SUPPRESSION, mg
WEIGHT, m©
SAMPLE SIZE
LOAD, a
TIME CONST.
dV, (mo'min)/in
dY, (10X ). frnils/min J/'n
Sets*
yj
;S^ ttjjvitry
fr
TEMPERATURE. *C ICHROMEL/ALUMEL)
Figure 3-3. DSC Scan of the MgS03 Solids Present in the 2.3M MgSO^ Solution

-------
wmwwi
wrfl fWT^Alt
OPBWn3R_^fS__	
SAMPLE: _ c„ 1M
aaa"'
ATM	*	
FLOW RATE	
T-AX1S
SCALE, :G/w~v _3Q_
PROG. RATE, C/man .
HEAT	COOI	ISO
SHIFT. In	+Sl__
DTA-PSC
SCALE. "C/in . lo
(mcal/sec)/fn ... .
WEIGHT. mflD.4 3
TGA
SCALE, mg/in		 .
SUPPRESSION*. mg_.
WEIGHT, mo	
TIME CONST, mc	
cfY\ f mo/min yin		
TMA
SCALE, m4s/in _
MODE	
SAMPLE SIZE.
LOAD, s	
cJY. C10X1. Cmte/lrninYin.
— I—




---i-
=1=
i
rt
Hi!
i •!
--'-V -

~ni+-
rm
-l-J .
rrti
_i
i ielis i


m.
£xvity

>
0
,.j:J
!i:i
i:.
;qri
1 —
•"-J" r

li:
+u:
in
i i -
I :I
	r—
-- :r:
—-i

'' i!
TTii
4-H i
;•)!
-«0 --too -SO -BO
in;
"=3Cr
~atr
~za~
-ea-
BO
10a
"iSrr
Tall
...j:
HI
TED	1BO
"SBo
TEMPERATLWE. "C (CHROMELVALUMEL)
Figure 3-4. DSC Scan of Pure MgSC>3*3H20

-------
PAtTWOW—I	
«JN qate3 ill
OPBWTOf^Mll	
SAMPLE Sfj
ATM	#_	
FLOW RATH	
T-AXIS
SCALE. "C/in	20
PHOG PATE, t/min
HEAT	COOI	ISO_
SHIFT, in	±£	
OTA-DSC
SCALE. -C/in _JO	
(nxa)/GecJ/«n_. ..
WEIGHT, rng _0 £	

TGA
SCALE, mg/in	
SUPPRESSION, mo_
WEIGHT, mo	
TffVTE CONST, mc^
tSY. Cmfl/rrMnVin	
TMA
SCALE. rni»s/in _
MODE		
SAMPLE SIZE-
LOAO. g	
dV. ( *3 X 3, C mUs/min )/tn .
-*SB

Sla tfcliEiftity
ft
I

TEMPERATURE, *C 1CHROMEL/ALUMEL)
Figure 3-5. DSC Scan of Pure MgS03*6H20

-------
run no. noon
RUN
SAMPLE- tV.S«VV<
^ !S*5»+r,*V
ftTMfc »f A i»-
FIJOW^*
T-AXIS
SCALE. C/m _ 20
PROG. BATE. •C/min	
HEAX_COOI	ISO-
SHIFT. >r>_
~ S
DTA-DSC
n
SCALE. "C/m _
(mcal/Mc}/in
WBGHT. mo^ii
nEFERENC^j^l^-jXa^
TGA
SCALE, mfl/m
suppression, r
WEIGHT, mo	
T*v*e CONST, OBC
dY. C mofrrin Vin	
TMA
SCALE. mi'sA'n.
MODE	
TTT
t-sj

SAMPLE SIZE.
LOAD, 9	
4a/minV«n -

$


TEMPERATUnE. "C (CHHOMEL/ALUMEL)
Figure 3<-6. DSC Scan of the Intermediate MgS03 Solids

-------
Figure 3-7. IR Speetra of Pure MgS03*3H20 and Pure MgS03*6H20

-------
Figure 3-8. IR Spectra of the MgS03 Solids Present in the 2.3M MgSOi, Solution and Pure MgS03*31120

-------
SECTION 4
CONCLUSIONS
The transition temperature of magnesium sulfite trihydrate
and hexahydrate phases can be lowered by increasing the con-
centration of the magnesium sulfate solution. It was observed
that at 37.5°C, MgS0 3*6H2 0 is the stable phase in a 1.5M mag-
nesium sulfate solution and MgS03*3Hz0 is the stable phase in a
2.3M magnesium sulfate solution. The theoretically predicted
equilibrium solution was approximately 2.0M MgSOi».
138

-------
TECHNICAL NOTE 200-045-36-04
EXPERIMENTAL RESULTS FOR THE
EQUILIBRIUM STUDIES ON MgSO* HYDRATES
Prepared by:
R. E. Pyle
139

-------
CONTENTS
1.	Introduction	143
2.	Experimental Approach for the Solubility
and Solubility Product Constant Studies	145
Experimental Apparatus 	 145
Experimental Procedures 	 146
Analytical Methods	147
3.	Experimental Results for the Solubility
and Solubility Product Constant Studies	148
Equilibrium Data Processing	148
Results for the Solubility and
Solubility Product Constant Studies .... 149
4.	Experimental Approach for the Ion-Pair
Dissociation Studies - Ion Selective
Electrode Measurements	157
Experimental Apparatus	157
Experimental Procedure 	 158
Analytical Methods	159
5.	Experimental Approach for the Ion-Pair
Dissociation Studies - Method of Influence
of Ion-Pair Formation on a Reference
Compound Solubility	161
Experimental Apparatus and Procedures . . , 161
Analytical Methods	162
6.	Experimental Results for the Ion-Pair
Dissociation Studies 	 165
140

-------
CONTENTS (Cont'd)
Equilibrium Data Processing - Ion
Selective Electrode Measurements 	 165
Equilibrium Data Processing - Method
of Influence of Ion-Pair Formation
on a Reference Compound Solubility 	 166
Results for the Ion-Pair Dissociation
Studies	166
7. Conclusions	170
Bibliography 	 171
FIGURES
Number Page
3-1 Solubility vs. Temperature for MgS03 Hydrates . . 153
3-2 Pure Solution Activity Product vs. Temperature. . 156
6-1	log1Q Kd vs. 1/T(K) x 103	167
141

-------
TABLES
Number	Page
3-1 Experimental Results for the MgS03 Solubility
Measurements	150
3-2 MgS03 Solubility Data	151
3-3 Experimental MgS03 Solubility Product Constants
(Ksp)	154
5-1	Results of Equilibrium Studies	 . 163
6-1	Ion-Pair Dissociation Constants 	 169
142

-------
SECTION 1
INTRODUCTION
The magnesium oxide or "Mag-Ox" process for sulfur dioxide
removal has proven to be technically feasible and is one of the
most advanced regenerative processes. In this scrubbing process,
the reaction of SO2 from flue gases with the MgO scrubbing slurry
yields the hydrated phases of MgS03 and other products. Experi-
ence has been that at a coal-fired unit with MgO scrubber the
hexahydrate crystalline phase was produced; and at an oil-fired
unit, the trihydrate crystalline phase was produced. From a
process point of view, the trihydrate phase may be more desir-
able because less energy is required to drive off the waters of
hydration in the regenerative cycle. However, the trihydrate
crystals that are produced are very small, on the order of a
few micrometers in size, while the hexahydrate crystals that are
produced are large massive particles several hundred micrometers
in size. Consequently, the settling and handling characteristics
of the two phases are vastly different. Currently the hexahydrate
phase is much easier to process in the solid separation equipment.
In view of these process problems, it was the objective of this
work to:
1)	investigate the theoretical aspects of hydrate
phase transitions,
2)	determine the equilibrium solubilities and solu-
bility product constants of the trihydrate and
hexahydrate phases of MgS0 3 as a function of
temperature,
3)	investigate the influence of solution composi-
tion on the transition temperature of the
hexahydrate-^trihydrate phase change, and
143

-------
4) investigate the influence of solution composi-
tion and temperature on the formation kinetics
of the two solid phases.
This Technical Note presents in detail the experimental
data that were derived from the equilibrium studies on the
hexahydrate and trihydrate crystalline phases of MgS03. The
equilibrium studies are separated into two phases of experimen-
tal work:
1)	the determination of the equilibrium solubili-
ties and the solubility product constants of
the trihydrate and hexahydrate phases of HgS03
as a function of temperature, and
2)	the determination of ion-pair dissociation
constants for the species: MSS0 3^aq^,
MgSOi^	and NaS°3~(aq) as a function of
temperature.
144

-------
SECTION 2
EXPERIMENTAL APPROACH FOR THE SOLUBILITY AND
SOLUBILITY PRODUCT CONSTANT STUDIES
EXPERIMENTAL APPARATUS
Solubility measurements for the MgS03 hydrates were per-
formed as a function of temperature over the range 25-65°C.
Excess magnesium sulfite hexahydrate or trihydrate crystals
were allowed to dissolve into approximately 300 ml of deoxy-
genated deionized water. This slurry was continuously agita-
ted by magnetic stirring in a closed reaction vessel specially
designed for these studies. The reaction vessel consisted of
a five-neck round-bottom flask of 500 ml capacity suspended in
an insulated constant temperature bath. With this arrange-
ment, the temperature of the reaction vessel could be held at
the desired reaction temperature to within +0.1°C.
Two necks of the multiple-neck reaction flask were used
as a continuous dry nitrogen purge gas inlet and outlet. The
dry nitrogen gas from the cylinder was passed through a deoxy-
genating scrubbing train consisting of two fritted 4-liter bot-
tles containing 1M Na2S03 prior to entering the sealed reaction
vessel. A fritted Na2S03 scrubbing bottle was also located at
the gas outlet. This nitrogen purge gas was used to minimize
oxidation of the sulfite ion in the reaction medium.
Another neck of the reaction vessel supported a ground
glass precision thermometer which penetrated the reaction solu-
tion. A fourth neck was used for periodic sampling which was
performed by forcing the slurry out of the reaction flask
through a 42mm, 0.8 micrometer Millipore filter membrane by
145

-------
pressurizing the vessel with the dry nitrogen purge gas. The
filtrate was then collected and the pertinent analyses per-
formed.
EXPERIMENTAL PROCEDURES
The clean reaction vessel was first charged with the de-
oxygenated deionized water and then heated to the desired re-
action temperature in the constant temperature bath. After
the continuous nitrogen purge gas stream was started, excess
MgS03 hydrate solids were added to the reaction vessel.
MgS03*6H20 crystals were used for the solubility determinations
below 40 °C and MgSO3»3H20 crystals were used for the solubility
determination above 40°C. The MgS0 3*6H20 crystals, 99% pure,
were obtained from Heico Laboratories. The MgS03 *31120 crys-
tals for these experiments were obtained by dehydrating
MgS03*6H20 crystals under their own atmosphere in a purged
oven at 100°C after the method of Dauerman (DA-001). The
equation for this reaction was:
MgS03-6H20(s)	MgS03'3H20(s) + 3H20(g). (2-1)
The trihydrate product was then analyzed for purity before use.
The dissolution reaction was allowed to come to equili-
brium while constant reaction temperature, continuous stirring
and nitrogen gas purging were maintained. The approach to
equilibrium was followed by periodically sampling the reaction
solution as described earlier. The solution samples were
analyzed for concentrations of magnesium, sulfite, and sul-
fate, solution pH, and reaction temperature (see Section 2.3
for Analytical Methods). The solubilities and solubility
146

-------
product constants were calculated from the equilibrium analyti-
cal data obtained at each reaction temperature.
ANALYTICAL METHODS
In this section, a brief outline of the analytical methods
used in the MgS03 hydrate equilibrium studies is given:
1)	Magnesium - colorimetric titration with Na2EDTA
or dilution preparation and atomic absorption
determination;
2)	Sulfite - iodine-buffer preparation and back
titration with standard arsenite or thiosulfate;
and
3)	Sulfate - peroxide oxidation of sulfite, hydrogen
ion exchange, and titration with standard sodium
hydroxide. Sulfate is calculated as the difference
between total sulfur and sulfite sulfur.
147

-------
SECTION 3
EXPERIMENTAL RESULTS FOR THE SOLUBILITY AND
SOLUBILITY PRODUCT CONSTANT STUDIES
EQUILIBRIUM DATA PROCESSING
Solubilities for MgS03.6H20 below 40°C and MgS03.3H20
above 40°C are determined from the experimental dissolution
analytical data. The solubilities in grams of MgSO3 per 100
grams of saturated solution are calculated from the equilibrium
concentrations of magnesium and/or sulfite in the saturated
solution at each experimental reaction temperature. That is:
S(g/100g) - [Ion] ^ • i	(3-1)
where:
S = solubility of MgS03 at temperature, T,
[Ion] = equilibrium magnesium or sulfite ion
concentration in moles/liter,
MW = molecular weight of MgS03, i.e. 104.37
grams/mole, and
D = density of MgS03 solution at the
particular reaction temperature in
grams/tnl (from literature data).
The solubility product constants for MgS03»6H20 and
MgS03*31120 are also determined from the experimental dissolu-
tion analytical data. The solubility product constants can
be expressed as:
148

-------
K
(T) = a
• a
• a
6
(3-2)
sp6
Mg+2 S03"2
H20
and,
• a
• a
(3-3)
SO 3"2
H20
Equilibrium activities for these particular solution species
are calculated by inputting pertinent reactor solution equili-
brium information, such as concentrations of magnesium and sul-
fite, solution pH, and reaction temperature, to the Radian
chemical equilibrium computer program. This program takes into
account the important and complex solution equilibria, such as
critical ion-pair formations, for each set of input reaction
conditions.
Reliable and accurate temperature-dependent dissociation
constant (K^) information for pertinent ion-pair species is
needed to calculate accurate solubility product constants from
experimental solubility data. Therefore, temperature dependent
ion-pair dissociation measurements for the species MgSS03(aq)»
MgS04£at^, and NaSOs-(aq) were performed using two different
experimental approaches. These ion-pair dissociation studies
are discussed in detail in Sections 4, 5 and 6.
RESULTS FOR THE SOLUBILITY AND SOLUBILITY PRODUCT
CONSTANT STUDIES
Experimental results for the solubility measurements on the
MgS0 3 system are summarized in Table 3-1. Experimental solu-
bilities for MgS0s*6Ha0 and MgS03»3H20, calculated from the
results in Table 3-1, are presented in Table 3-2 along with
solubility data taken from the literature (LI-001).
149

-------
TABLE 3-1. EXPERIMENTAL	RESULTS FOR THE MgS03	SOLUBILITY MEASUREMENTS
Run	Solution	Reactor Temp.	Cone, of Magnesium	Cone, of Sulfite	Cone, of Sulfate
No.	pH	(°C)	(mole/5.)	(mole/£.)	(mole/£)
1	9.54	25.0	0.0638	0.0631	0.00098
2	9.39	30.0	0.069	0.061	0.0094
3	9.30	35.0	0.082	0.075	0.0082
4	8.51	45.0	0.0816	0.086	0.000
5	8.47	55.0	0.0769	0.077	0.000
6	8.32	65.0	0.067	0.067	0.002

-------
TABLE 3-2. MgS03 SOLUBILITY DATA
Temperature	Stable	Experimental Solubility	Literature Solubility
(8C)	Solid Phase	(g MgS03/100 g)	(g MgS03/100 g)
25.0	MgS03-6H20	0.668	0.625
30.0	MgS03*6H20	0.723	0.721
35.0	MgS03*6H20	0.861	0.835
45.0	MgS03*3H20	0.906	0.904
55.0	MgS03 *3H20	0.814	0.790
65.0	MgS03*3H20	0.713	0.710

-------
In Figure 3-1, a plot of the MgS03 solubility data from the
literature (LI-100) versus temperature is presented along with
our experimentally derived values for comparison.
Solubility product constants for MgS03«6H20 and MgS03.3H20
derived from the literature solubility data are presented in
tabular form as a function of temperature in Table 3-3.
It is also convenient to plot the solubility tendencies
of the hexahydrate species on the same graph as the trihydrate
species. As defined earlier, the pure solution activity pro-
ducts or solubility product constants for the hydrate species
can be written as:
It is clear that the hexahydrate activity product has units
that differ from the trihydrate activity product units. There-
fore, if we consider the ordinate on our ap3 plot as y, then
we can plot ap6 values as a function of temperature on our ap3
graph by defining the y ordinate as:
KSpt(T) - 
-------
2.0
2.0
1.8
1.8
1.6
1.6
1.4
1.4
Literature
1.2
1.2
Our Experimental
Data
1.0
1.0
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
0
40
10
20
30
50
70
80
0
90 100 110
(T"C)
Figure 3-1. Solubility vs. Temperature for MgS03 Hydrates

-------
TABLE 3-3. EXPERIMENTAL MgS03 SOLUBILITY PRODUCT CONSTANTS (Ksp)
Temperature
(°C)
Stable
Solid Phase
Ksp

0
MgS03"6H20

4.537 x
10" 5
10
MgSO 3•6H20

4.797 x
10" 5
20
MgSO 3-6H20

5.368 x
10" 5
25
MgSO 3•6H20

5.720 x
10-5
30
MgSO 3•6H 20

6.124 x
10" 5
35
MgSO 3•6H 20

6.579 x
10-5
40
MgSO 3•6H 20

7.120 x
10"5
45
MgSO 3¦6H20
(metastable)
7.735 x
10" s
50
MgSO 3•6H 20
(metastable)
8.375 x
10" s
55
MgSO 3•6H20
(metastable)
9.060 x
10" 5
60
MgSO 3•6H2O
(metastable)
9.786 x
10"5
62.5
MgSO 3•6H20
(metastable)
1.013 x
10-*
30
MgS03-3H20
(metastable)
1.074 x
10-*
35
MgSO 3-3H20
(metastable)
8.662 x
10" 5
40
MgSO 3•3H20
(stable)
7.120 x
10-5
45
MgSO3•3H20
(stable)
6.074 x
10"5
50
MgSO 3•3H20
(stable)
5.191 x
10"5
55
MgSO 3•3H20
(stable)
4.466 x
10-5
60
MgSO 3 * 3H20
(stable)
3.894 x
10~s
70
MgSO 3•3H20
(stable)
3.020 x
10~5
80
MgSO 3¦3H20
(stable)
2.457 x
10-5
90
MgSO 3•3H20
(stable)
2.059 x
10-5
100
MgSO 3•3H20
(stable)
1.771 x
10~5
154

-------
Therefore, in Figure 3-2, a plot of y versus temperature is
presented. The activity products for the solubility product
constants were calculated using the Radian chemical equilibrium
computer program as described earlier. No solubility product
constants for the MgS03 hydrates were found in the literature
to compare with our experimentally derived values.
155

-------
u>
O
X
5s
O
N
V
C4
|«o
o
CO
CM
I?
11.0
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
HgSO3«3H20
11.0
10.0
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0 10 20 30 40 50 60 70 80 90 100 110 120
(T°C)
0
Figure 3-2. Pure Solution Activity Product vs. Temperature

-------
SECTION 4
EXPERIMENTAL APPROACH FOR THE ION-PAIR DISSOCIATION
STUDIES - ION SELECTIVE ELECTRODE MEASUREMENTS
EXPERIMENTAL APPARATUS
Ion-pair dissociation measurements for the species
MgS03^a(^, MgSO^^^, and NaS03-^a(^ were performed as a func-
tion of temperature over the range 25-55°C. Unsaturated test
solutions of the species of interest were prepared and brought
to the desired experimental temperature. The free cation re-
sponse of the solution was then measured with special cation
selective electrodes. This value was compared to the measured
free cation response of standard solutions with negligible
ion pairing.
Test solutions for the particular ion-pair species were
prepared from either MgS03*6H20, MgSO^, or Na2S03 by dissolu-
tion of the respective pure solid in approximately 150 ml of
deoxygenated deionized water to a predetermined unsaturated
concentration with pH adjustment if necessary. The test solu-
tions were continuously agitated by magnetic stirring in closed
reaction vessels used specifically for these equilibrium stud-
ies. These vessels were Nalge high temperature, wide-mouth,
flat-bottom plastic bottles with sealable tops and of 250-ml
capacity. The sealed reaction vessels containing the ap-
propriate test and cation standard solutions for each experi-
ment were supported in an insulated constant temperature bath
that maintained the desired reaction temperature to within
±0.1°C.
A 4-holed rubber stopper was used to support a precision
thermometer for accurate solution temperature measurements,
157

-------
an input divalent or sodium cation electrode for free cation
activity measurements, an Orion single junction or Beckman
standard saturated calomel reference electrode, and a standard
pH electrode for solution pH measurements. The thermometer-
electrode assembly was inserted consecutively in the appro-
priate standard and test solutions maintained at the desired
experimental temperature. The rubber stopper provided an air-
tight seal with the Nalge reaction vessels to minimize oxida-
tion of sulfite during the measurements. In addition, a
nitrogen blanket for the solutions was also provided. The
solution pH and free cation response measurements were taken
with a calibrated Model 701 Orion Digital Meter.
EXPERIMENTAL PROCEDURE
A series of standard solutions of different concentrations
with negligible ion pairing were prepared quantitatively. Un-
saturated magnesium ion standard solutions were prepared from
pure MgCl2*6H20, and sodium ion standard solutions from pure
NaCl, Sodium and magnesium concentrations of the standard
solutions were determined accurately with dilution preparation
and atomic absorption spectroscopy. Concentrations of the
standard solutions ranged over approximately two decades of
molarity with the extremes' bracketing the concentrations of
the particular ion-pair test solutions under investigation.
The sodium or magnesium standard solutions contained in
the sealed Nalge reaction vessels were then placed in the con-
stant temperature bath maintained at 25°C and allowed to therm-
ally equilibrate. The solution temperature, pH, and free ca-
tion response of each standard solution was then measured in
duplicate by the technique described in Section 4.1.
158

-------
Equilibrium activities for the particular standard solu-
tion species were calculated by inputting the ion concentra-
tions, solution pH, and temperature for each standard solution
into the Radian chemical equilibrium computer program. Cali-
bration curves for the cation selective electrodes were pre-
pared by plotting the logarithm of the sodium or magnesium
free-cation calculated activity versus the measured electrode
response in millivolts for each standard solution.
Test solutions of different concentrations and of the
desired ion-pairs were then prepared in duplicate for each
experimental measurement and analyzed for magnesium, sulfite,
sulfate, or sodium before and after the response measurement at
each experimental temperature. The thermal and drift corrected
free-cation response derived from the measurements of these
test solutions at the various experimental temperatures were
then compared to the free-cation electrode response from the
calibration curve for the particular standard solutions.
Activities for the particular test solution species were cal-
culated, as before, by using the Radian chemical equilibrium
computer program. From these experimental measurements and
computer calculations, temperature-dependent dissociation con-
stants (Kj) for the particular ion-pairs were calculated
(see Section 6) .
ANALYTICAL METHODS
The following analytical methods were used in the ion-
selective electrode ion-pair dissociation studies:
1) Magnesium - colorimetric titration with Na2EDTA
or dilution preparation and atomic absorption
determination.
159

-------
2)	Sodium - dilution preparation and atomic ab-
sorption determination.
3)	Sulfite - iodine-buffer preparation and back
titration with standard arsenite or thio-
sulfate.
4)	Sulfate - peroxide oxidation of sulfite, hy-
drogen ion exchange, and titration with stand-
ard sodium hydroxide. Sulfate is calculated as
the difference between total sulfur and sul-
fite sulfur.
160

-------
SECTION 5
EXPERIMENTAL APPROACH FOR THE ION-PAIR DISSOCIATION
STUDIES - METHOD OF INFLUENCE OF ION-PAIR FORMATION
ON A REFERENCE COMPOUND SOLUBILITY
The measurement of the solubility of a sparingly soluble
reference compound in the presence of another electrolyte may
also be used to derive dissociation constants if the equili-
brium constants for the reference compounds are known. The
solubility of the reference compound is increased by an elec-
trolyte containing ions which can form ion-pairs with the
saturating salt. For example, the solubility of CaS03*%H20
is increased in the presence of MgCl2, primarily due to the
following reaction:
The equilibrium constant of reaction 5-lb can be derived from
a complete chemical analysis of the equilibrium solution and
calculations which consider other pertinent reactions and
activity coefficient effects. For these calculations, the
Radian chemical equilibrium computer program is used.
EXPERIMENTAL APPARATUS AND PROCEDURE
In this work CaS0s«%H20 and CaS0t»*2H20 were used as the
sparingly soluble reference compounds. The salts equilibrated
with each solid in individual runs were MgCl2, NaCl, and CaCl2.
Chloride salts were used since the chloride anion does not
+ 9	+ 9	+
form ion-pairs with Mg , Ca , or Na to any appreciable
Mg+I + so? ; MgS03(aq)
(5-la)
with
Kd = aMg+2 ' aS0s2/a:
MgSO 3(aq).
(5-lb)
161

-------
extent. Experiments with CaCl2 were done in order to check
the reference compound equilibrium constants.
The added electrolyte concentration was varied from 0.01 M
to 0.6 M depending on the particular salt used and selected run
condition. Controlled experimental reaction temperatures of
30°C and 50°C were used. The equilibrium experiments using
freshly prepared solid were carried out in sealed, cylindrical
tumblers with reaction times varying from 3-5 days. At the
end of each run, solution pH was measured and samples were
withdrawn for analysis. The results of these analyses are
presented in Table 5-1.
ANALYTICAL METHODS
The following analytical methods were used in the ion-
pair dissociation studies:
1)	Magnesium and calcium - colorimetric titration
with Na2EDTA or dilution preparation and atomic
absorption determination.
2)	Sodium - dilution preparation and atomic absorp-
tion determination.
3)	Sulfite - iodine-buffer preparation and back
titration with standard arsenite or thiosulfate.
4)	Sulfate - peroxide oxidation of sulfite, hydro-
gen ion exchange, and titration with standard
sodium hydroxide. Sulfate is calculated as the
difference between total sulfur and sulfite sulfur.
162

-------
TABLE 5-1. RESULTS OF EQUILIBRIUM STUDIES
Run Condition*	Analytical Raaulta (acolaa/lltar)
Date
Run
No.
lap.
CC)
Solids
Llsuor
dH
Tamp.
CO
Ca**

Na+
CI"
so'
sol
2/19/76
13
50
CaSOi'JfiUO
D.l. H,0
8.20
48.0
1.09

	
	
0.39
0.42
II
14
"
"
0.2 M MgClj
8.63
48.8
5.13
198
	
399
3.89
1.57
«t
15

"
0.6 H Nad
9.81
46.6
2.26
	
588
588
2.26
.57
t>
16
II
CaSOi,'2H,D
D.I. HjO
3.50
48.5
14.9
	
	
	
	
15.1
II
17
II
"
0.2 H MgCla
8.49
48.0
43.4
193
	
395
	
43.2

18
It
¦¦
0.6 M KaCl
9.3*
48.0
37.9
—
572
580
	
37.4
2/24/76
19
30
CaSOi'VHsO
D.l. HaO
7.90
30.0
1.01
...
...
—
0.67
0.17
"
20
It
II
0.2 M J"»Clj
8.56
30.5
5.11
lf.3
	
3'5
4.54
.50
II
21
••

0.6 M NaCl
8.49
30.8
3.08
	
529
524
2.39
.64
It
22
"
II
0.1 M MgCls
8.03
30.0
4.04
108
	
213
3.33
.76
II
23
M
H
0.3 M HaCl
8.72
30.0
2.45
	
2B5
292
1.94
.75
It
24
"
11
0.2 H CaCli
8.85
30.0
177.
	
	
354
0.25
.28
3/04/76
25
30
CiSOw* 2H]0
D.l. H»0
7.51 .
33.0
14.6
—
	
	
	
H.5
II
26
it

0.2 H M|Cli
7.97
33.4
40.5
223
	
402
	
41.3
It
27
t«
"
0.6 M NaCl
8.20
33.0
34.5
	
569
581
	
35.4
II
28
ii
M
0.1 M MgCl]
8.02
33.0
31.6
135
	
217
	
33.7
• 1
29
H
It
0.3 M NaCl
8.38
33.0
27.4
—
278
286
	
29.2
II
30
II
M
0.2 K CaClj
8.81
33.0
200.
—
	
364
	
8.31
3/09/76
31
30
CaSOi-^HtO
D.l. HjO
7.51
32.0
1.11
—
	
	
0.60
0.85
II
32
II
It
0.01 M CaClj
7.10
32.0
11.5
—
	
20.4
0.26
0.67
II
33
"
fl
0.1 M CaClt
7.09
31.5
92.8
—
	
186
0.18
0.49
3/13/76
34*
SO
C«S01 *%H >0'
D.l. H,0
6.23
50.5
0.99
—
...

0.54
0.63
II
33*
*1

0.6 M NaCl
6.14
50.3
2.52
—
600
580
2.08
2.21
"
36*
II
II
0.2 M MgCli
6.16
50.5
4.73
208
	
405
4.09
4.10
II
37*
I*
H
0.2 M CaCla
6.08
50.3
200.
	-
	
400.
0.25
0.32
* pH adjuatad Co 6 with HC1.
163

-------
5) Chloride - potentiometric titration with stan-
dard silver nitrate.
164

-------
SECTION 6
EXPERIMENTAL RESULTS FOR THE ION-PAIR DISSOCIATION STUDIES
EQUILIBRIUM DATA PROCESSING - ION SELECTIVE ELECTRODE
MEASUREMENTS
Given an ion pair such as	with the following equi-
librium dissociation in solution:
MA(aq) t M+2 + A*"2	(6-1)
an equilibrium dissociation constant for this reaction can be
written as:
aM+2 • a.-2
Kd = 	*—	(6-2)
M(aq)
where a is the activity for the particular solution species
and is a function of temperature. The activity of a par-
ticular solution species is related to its solution concentra-
tion by the relationship:
a = y * M	(6-3)
where
Y == the activity coefficient, and
M - the molality of the species.
At temperature T the dissociation constant for a particu-
lar ion-pair is calculated by an iterative trial and error
procedure using the measured experimental data (LO-OOl). A
trial value for and the experimentally determined total
165

-------
concentrations of all solution species are input to the Radian
chemical equilibrium computer program. The computer-calculated
free cation activity is then compared to the value derived from
the cation selective electrode measurements of the appropriate
test solutions. The accepted value of at temperature T is
the value that minimizes the difference between the experimen-
tally and theoretically derived values of activity for the free
cation of interest.
EQUILIBRIUM DATA PROCESSING - METHOD OF INFLUENCE OF ION-PAIR
FORMATION ON A REFERENCE COMPOUND SOLUBILITY
The pH, temperature, and chemical analysis results from
these equilibrium experiments are input to the Radian chemical
equilibrium computer program. For each run a series of equi-
librium distributions and relative saturations are calculated
as a function of the input dissociation constants for the
reaction considered. The actual dissociation constant is then
obtained graphically as that value which yields a calculated
relative saturation of 1.00 for the particular sparingly solu-
ble reference compound. Effects of probable errors in the chem-
ical analyses are estimated by varying the input concentrations
over the expected range and noting the change in calculated
dissociation constant.
RESULTS FOR THE ION-PAIR DISSOCIATION STUDIES
The dissociation constants for MgS03 , MgSOi* ^ , and
NaSOl^ ^ were determined at 25, 35, 45 and 55°C by the ion
specific electrode method. The same constants were also
determined at 30 and 50°C by the sparingly soluble salt
method. The sparingly soluble salt data are plotted as
logioK^ versus 1/T(K) in Figure 6-1. This figure is used to
166

-------
-4.oor
-3.00
-2.00
-1.00
50°C
e
-e
MgSO 3
(aq)
MgSO 4
(aq)
NaS03
(aq)
Experimental
Data
e
3.00 3.05
3,10 3,15 3.20 3.25
1/T(K) x 103
3.30 3.35 3.40
Figure 6-1. ^°8^q vs* l/T(K) x 10®
167

-------
extrapolate and interpolate points at 25, 35, 45, and 55°C.
These points and the experimental points for the ion specific
electrode method are given in Table 6-1.
168

-------
TABLE 6-1. ION-PAIR DISSOCIATION CONSTANTS
K dissociation
Sparingly
Ion-Specific Soluble
Ion-Pair	Temperature (°C)	Electrode	Compound
MgS03
(aq)
25
35
45
55
MgSOi*
(aq)
25
35
45
55
NaSO;
(aq)
25
35
45
55
0.00173
0.00140
0.00113
0.00091
0.00136
0.00110
0.000896
0.000747
0.00347
0.00287
0.00238
0.00196
0.00569
0.00533
0.00501
0.00473
0.055
0.055
0.055
0.055
0.06611
0.05760
0.05063
0.04484
X69

-------
SECTION 7
CONCLUSIONS
It was found that the experimental solubilities of the
MgS0 3 hydrate solids determined by Radian agree quite well
with the literature values (LInOOl), From these temperature-
dependent solubilities accurate solubility product constants
were calculated by using the Radian experimentally determined
dissociation constants. It was felt that in order to arrive
at accurate temperature-dependent solubility product constants,
an update of critical equilibrium dissociation constants was
needed. Therefore, a great deal of effort was spent in deter-
mining accurate temperature-dependent dissociation constants
for MgS03(;aq) , MgSO„£ j andNaS03"( ^ ion-pairs.
It was found that the method of influence of ion-pair
formation on the solubility of a sparingly soluble reference
compound to determine ion-pair dissociation constants yielded
generally more reliable results than the ion-selective elec-
trode experiments. This was because the selective cation
response measurements were found to be offset by a non-negli-
gible electrode junction potential. This electrode junction
potential, inherent in every measurement, is a function of
test solution composition and temperature and is of unknown
magnitude. Therefore, the temperature-dependent dissociation
constants from Table 6-2 (specifically MgSOj^ were used
in calculating accurate temperature-dependent solubility
product constants for the HgS03 hydrate solids.
170

-------
BIBLIOGRAPHY
DA-001 Dauerman, Leonard, The Thermal Dehydration of
MgSQa*6H20 and MgSOa'SHaO, Chemical Construction
Company, 1975.
LI-001 Link, W. F., Solubilities of Inorganic and Metal
Organic Compounds - Vol. II, Washington, D. C.,
American Chemical Society, p. 985.
LO-OOl Lowell, P. S., et al., A Theoretical Description of
the Limestone Injection-Wet Scrubbing Process,
Contract No. CPA-22-69-138, Austin, Texas; Radian
Corporation, 1970.
171

-------
TECHNICAL NOTE 200-045-36-05
EXPERIMENTAL RESULTS FOR PRECIPITATION
KINETICS STUDIES ON MgS03 HYDRATES
Prepared by:
R. E. Pyle
172

-------
CONTENTS
1.	Introduction	174
2.	Experimental Approach	
Experimental Apparatus and Procedure	176
Experimental Sampling Procedure 	 176
Analytical Methods	177
3.	Experimental Results
Kinetics Data Processing	179
Results	184
4.	Discussion of Results	189
Bibliography
Appendix
FIGURES
Number	Page
3-1 MgS03»6H20 Precipitation Rate vs. Relative
Saturation at Various Temperatures 	 185
3-2 MgS03*3H20 Precipitation Rate vs. Relative
Saturation at 55°C	 186
3-3 MgS09*3H20 Precipitation Rate vs. Relative
Saturation at 70°C	 187
3-4 MgS03*3H20 Precipitation Rate vs. Relative
Saturation at 85°C		 188
TABLES
Number Pa%e
3-1 . .	 193
173

-------
SECTION 1
INTRODUCTION
In an MgO scrubber system, a circulating slurry of MgO and
hydrated MgS0 3 is used to absorb S02 from a. flue gas stream.
The S02 is removed from the scrubber system as a precipitate of
hydrated magnesium sulfite. Two hydrates have been identified,
MgSO3•3H20 and MgS03*6H20. The crystals formed in the instal-
lation removing S02 from the flue gases of Boston Edison's
oil-fired boiler were found to be primarily trihydrate crystals.
The installation at PEPCO's coal-fired boiler produced primarily
hexahydrate crystals.
The removal of S02 from flue gas streams by an MgO regen-
erative scrubbing system has proven to be technically feasible.
The status of technology at the time the above units were designed
was insufficient to predict which hydrates form would be favored.
An understanding of the phenomena involved in the. production of
either hydrated form is important. The trihydrate crystals that
are produced are very small in size. In comparison,, the settling
and handling characteristics of these two hydrated phases are
markedly different. From a process point of view, the trihydrate
may be the more desirable crystalline form because it takes less
energy to drive off the waters of hydration when the crystals
are calcined to regenerate the MgO. The economics may depend,
however, on the form of available heat. It is presently easier
from an operating standpoint to produce hexahydrate crystals
because of the ease with which the hexahydrate solids can be
removed from the liquids in the scrubber.
In view of these process problems , it is clear that the
chemistry involved in the preferential production of either
hydrated phase of MgS03 must be characterized. It was, therefore,
the objectives of this work to:
174

-------
(1)	investigate the theoretical aspects of hydrate phase
transitions,
(2)	to determine the equilibrium solubilities and
solubility product constants of the trihydrate and
hexahydrate phases of MgS03 as a function of
temperature,
(3)	investigate the influence of solution composition
on the transition temperature of the hexahydrate-
trihydrate phase change, and
(4)	investigate the influence of solution composition
and temperature on the formation kinetics of the
two solid phases.
This Technical Note presents in detail the experimental
results that were derived from the precipitation kinetics studies
on the hexahydrate and trihydrate crystalline phases of MgS03.
Experimental details of the equilibrium studies may be found in
Technical Note No. 200-045-36-04, and results of the transition
temperature lowering studies may be found in Technical Notes
No. 200-045-36-02 and No. 200-045-36-03.
175

-------
SECTION 2
EXPERIMENTAL APPROACH
EXPERIMENTAL APPARATUS AND PROCEDURE
The precipitation kinetics experiments on MgS03*6H20 and
MgS03'3H20 were performed by using batch-solid/batch-liquid
reaction techniques. It was found that this experimental
technique works well for these more soluble salts although data
analysis is generally more difficult. For low reaction tempera-
tures (<50°C), a batch-solid/batch-liquid crystallizer of
plexiglass construction and an external stirring motor were used
for the kinetics experiments. For higher reaction temperatures,
a pyrex reactor of approximately 3-liter capacity was used.
Supersaturated solutions were produced in the well-stirred
reactors by introducing equal volumes of the two separate feed-
stock solutions, especially MgGl2 and Na2S03 of predetermined
concentrations. The reactor was situated in a constant tempera-
ture bath held at the desired experimental reaction temperature.
Within the reactor, precipitation from the supersaturated solution
was initiated by the addition of MgS03*6H20 or MgS03 3H20 seed
drystals. From this point, the reactor contents could be sampled
and the various analytical determinations performed.
EXPERIMENTAL SAMPLING PROCEDURES
Initial samples of the mixed feedstock solutions, speci-
fically MgCl2 and Na2S03, were taken from the reactor contents
just prior to adding seed crystals. Determination of magnesium
and sulfite was performed on these samples to establish the
initial concentrations of the reacting species. In addition,
solution pH and temperature were taken.
176

-------
At predetermined intervals during an experimental run,
samples of the reactor contents were taken until the crystal
growth reaction was complete. Determinations of magnesium,
sulfite and total sulfur were performed on the liquid phase
samples. Total sulfur determinations were used to determine the
extent of sulfite oxidation to sulfate. (It was found that
sulfite oxidation was negligible in these high sulfite concen-
tration solutions.) In addition, weight percent solids,
solution pH, and temperature were taken. In this way, the re-
action progress could be monitored regularly throughout the
experimental run.
The results of these analytical determinations were primarily
used in quantifying the two principal experimental quantities of
interest, namely: the rate of precipitation and the reaction
solution relative supersaturation.
ANALYTICAL METHODS
In this section, a brief outline of the analytical methods
used in the MgSOa hexahydrate and trihydrate precipitation
kinetics studies is given.
(1)	Magnesium - colorimetric titration with Na EDTA
or dilution preparation and atomic absorption
determination;
(2)	Sodium - dilution preparation and atomic absorption
determination;
(3)	Chloride - silver nitrate potentiometric titration;
(4)	Sulfite - iodine-buffer preparation and back
titration with arsenite; and
177

-------
Total Sulfur - peroxide oxidation of sulfite,
hydrogen ion exchange, and titration with
standard sodium hydroxide.
178

-------
SECTION 3
EXPERIMENTAL RESULTS
KINETICS DATA PROCESSING
Precipitation Rate
For the batch-solid/batch-liquid kinetics experiments, a
precipitation rate can be calculated from the amount of solid
material produced as a function of the time. The slope of the
total amount of solids versus reaction time curve is equal to
the precipitation rate or rate of growth for the precipitating
species. Therefore, the precipitation rate at time t can be
expressed as:
R = lim(l/MW) (~)
At+o
t<
= (1/MW)(5|)
=t	ar t=t
o	o
(3-1)
where,
N » mass of solids produced in grams,
t - time duration of the run in min,
MW = molecular weight of the precipitating species
in grams/mMble, and
R ¦ precipitation rate in mMoles/min at time t .
The total mass of solids, N, was determined from the weight
percent solids measurements or solution material balances at
predetermined time intervals throughout the kinetics experiment.
179

-------
Equation 3-1 expresses the precipitation rate at time t
or the rate at which solid material is being produced in the
reactor at time t . The experimental precipitation rate is
determined then by approximating Equation 3-1 and manually cal-
culating slopes of the amount of product versus reaction time
curve at various times throughout the kinetics experiment.
The amount of precipitating material produced at time t
can be determined not only through weight percent solids measure-
ments, but also from material balance criteria. That is, the
change in the solution magnesium or sulfite ion concentration
with time determined by analytical means must be equivalent to
the number of moles of magnesium or sulfite reacted to form the
MgS03 hydrate product. Therefore, the amount of precipitated
material produced at time t can be calculated from the following
equation:
N(t ) = ([Ion] - [Ion]	)-Vol	-MW (3-2)
O	Reactor
where,
N(t ) = mass of MgS03 hydrate product material
at time t in grams,
[Ion]t = initial magnesium or sulfite ion concen-
tration in solution in moles/l,
[Ion].^ = magnesium or sulfite ion concentration in
o
solution at time t in moles/I,
o	'
dL,	= total reaction solution volume in liters,
Tleactor
and
180

-------
MW = molecular weight of precipitating MgS03
hydrate product in g/mole.
This determination provides an excellent method in addition to
the weight percent solids measurements for calculating the pre-
cipitation rates.
Solution Relative Supersaturation
The MgS03 hydrate precipitation rate, R, at time t
determined in the manner previously described, is a function of
the solution relative supersaturation. The relative supersatura-
tion is defined as the ratio of the reaction solution activity
product for the precipitating species to the equilibrium solu-
bility product, K , for the precipitating species. In this
sp
case:
R.S., - aMg«-»S0--4!0/K»PMgS0j.6H!0	(3_3)
or
K-S.j - aMg+2**S0;2,aH20^K«PMgS03.3H20
Activities for the particular solution species are calculated by
inputting pertinent reaction solution information for reaction
time t such as concentrations of magnesium, sodium, chloride,
o
sulfate and sulfite, pH, and temperature, to the Radian chemical
equilibrium computer program.
The Rate Equation
A suitable rate expression for the MgS03 hydrate solid
precipitation from supersaturated liquor may be written in the
181

-------
following form-.
R = k • M • 4>	(3-5)
where,
R = rate of solid precipitation,
k = rate constant, which may vary with liquor
temperature, composition, and transport
parameters,
 = driving force which is related to the degree
of MgS03 supersaturation, and
M = term dependent on the amount of solid phase
present.
The term M is often assumed to be proportional to the
exposed surface area of the solid phase. This is difficult to
quantify in experiments with suspensions of many fine particles
of seed crystals. Consequently, no crystal surface area
measurements were attempted. However, MgS03 hexahydrate crys-
tals, previously grown from solution and analyzed for purity,
were ground and then sieved through 400 mesh wire screen. The
resulting sieved product, with an average particle diameter of
less than or equal to 37 microns, was used as uniform seed
material for all MgS03 hexahydrate kinetics experiments in order
to facilitate rate data correlation. The measured precipitation
rates were, subsequently, normalized by dividing by the mass of
crystals at time t .
For dissolution and precipitation reactions, the driving
force term, 
-------
the actual and equilibrium quantities of the reacting species.
A general form for the driving force function (NA-033) can
therefore be written as:
4> - [Up)1/n- K "V	(3-6)
8P
Where
ap = the activity product for the
precipitating species, i.e.
aMg+2 ' aS03"2 ' a6H20
or
^g"1"2 ' aS03-2 ' a3H20'
Kgp = the equilibrium solubility product
constant for the precipitating
species,
n = an exponent generally taken to be
equal to the number of cations plus
anions, in this case n - 2.
Equation 3-6 may also be written in the following form:
~ - KaD rR.s.1/n - 1]°	(3-7)
sp
where
R.S. ¦ the solution relative supersaturation for the
precipitating species.
With this general form for the driving force, the expression for
183

-------
the rate of precipitation (3-5) can now be written as:
R(mMoles/min) = k(mMoles/gram-min) •M(grams)	(3-8)
• KsptR.S. 1/n - l]n
In this kinetics study, the precipitation rate normalized by
dividing by the mass term, M, is analyzed as a function of the
solution relative supersaturation.
RESULTS
Experimental results for the precipitation kinetics studies
on the MgS03 system have been summarized in Table 3-1, and are
presented in the Appendix. The relative saturations were cal-
culated using the Radian chemical equilibrium computer program
as described earlier. The precipitation rates for MgS03»6H20
and MgS03*3H20 (in mMoles/gram-min) at various experimental
temperatures are plotted versus solution relative saturation in
Figures 3-1 through 3-4. The reported precipitation rates were
calculated by using the methods described in Kinetics Data
Processing.
184

-------
(30°)
1.
0.
0.
0.
(20°)
0.
0.
0.
0
0.
0.
0.
1.50
Relative Saturation
2.50
00
2.00
Figure 3-1. MgS03*6H20 Precipitation Rate vs
Relative Saturation at Various Temperatures
185

-------
0.0030 t
0.0025
0.0020
0.0015
0.0010
0.0005
0.0
1.00
1.50
2.00
Relative Saturation
2.50
Figure 3-2. MgSOa^HaO Precipitation
Rate vs Relative Saturation at 55°C
186

-------
0.12 ,
0.11 ¦
0.10
0.09 -
0.08
0.07 ¦
0.06 ¦¦
0.05
0.04 ..
O.OJ
0.02 ••
0.01
0.0
1.00
Relative Saturation
Figure 3-3. MgS0s*3H20 Precipitation Rate vs Relative
Saturation at 70®C
187

-------
0.12
0.11
0.10
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02 .
0.01 .
0.0
1.50
2.00
1.00
2.50
Relative Saturation
Figure 3-4. MgS03*3H20 Precipitation Rate vs
Relative Saturation at 85°C
188

-------
SECTION 4
DISCUSSION OF RESULTS
As stated in Section 1.0, experience has been that at a
coal-fired unit with MgO scrubber the hexahydrate crystalline
phase was produced, although scrubber operating temperatures of
near 55°C would suggest the thermodynamically favored trihydrate
product. Also, Technical Notes 200-045-36-02 and 200-045-36-03
demonstrate both theoretically (Technical Note 200-045-36-02) and
experimentally (Technical Note 200-045-36-03) that the phase
transition temperature between MgS03-6H20 and MgS03»3H2Q solid
in aqueous solution is dependent upon the solution composition.
Furthermore, the transition temperature is lowered from a maxi-
mum of 41°C as the dissolved solids content of the solution
increases. Therefore, the precipitation of MgS0j*6H20 at
scrubber conditions cannot be explained from an equilibrium
point of view but must be the result of a kinetic phenomenon.
Indeed, the experimental results of this Technical Note tend
to verify this assumption.
As can be seen from Figures 3-1 through 3-4, there is a
wide variation in growth rate magnitudes between the hexahydrate
and trihydrate phases and over the range of temperature studied.
For example, the hexahydrate precipitation rate at 30°C is
nearly three orders of magnitude greater than the trihydrate
precipitation rate at 55 °C at the same relative saturation or
driving force. In fact, the trihydrate precipitation rate only
begins to approach magnitudes characteristics of hexahydrate
growth near reaction temperatures of 85°C. This temperature
enhancement of the rate is of course expected from the Arrhenius
equation. Nevertheless, the growth rates of the two hydrate
phases near scrubber operating temperatures are sufficiently
disparate that hexahydrate precipitation under these conditions
189

-------
must be the result of kinetic factors. Kinetic effects under
conditions more typical of actual operating scrubber conditions
are being investigated in Task 54.
190

-------
BIBLIOGRAPHY
NA-033 Nancollas, George H. and N. Purdie, "The Kinetics of
Crystal Growth", Chem. Soc. Quarterly Rev., 18, 1-20
(1964).
191

-------
APPENDIX
TO TECHNICAL NOTE
200-045-36-05
192

-------
TABLE 3-1
RUN NO. 1
MgS0j-6HjO 0 20°C
(Hin.)
Ti»e/Date	Temp °C	gH	(Hg++)	(Na+)
t=0 tl/18/75	22	9.10	.095	.188
t-10 11/18/75	22	9.15	.090	.188
t=20 11/18/75	22	9.20	.087	.188
t=30 11/18/75	22	9.40	.086	.188
t=40 11/18/75	21.5	9.25	.084	.188
t=50 11/18/75	21.5	9.20	.082	.188
t=60 11/18/75	21	9.20	.083	.188
t=120 11/18/75	20.9	9.23	.077	.188
t=240 11/18/75	20	9.20	.083	.188
t=300 11/18/75	20	9.10	.090	.188
(CI")

-------
TABLE 3-1 (Cont'd)
RUN NO. 2
MgS03-6H20 ? 20<5C
(Hin.)
Ti«e/Date
Temp °C
PH
(Mg++)
(Na+)
t=0 11/20/75
21
9.1
.121
.235
t=10 11/20/75
21
9.1
.100
. 235
t»20 11/20/75
21
9.1
.092
.235
t=30 11/20/75
21
9.1
.085
.235
t-40 11/20/75
21
9.1
.083
.235
t-50 11/20/75
20
9.1
.085
.235
t=60 11/20/75
20
9.1
.084
.235
t=120 11/20/75
20
9.1
.083
.235
t=240 11/20/75
20.5
9.1
.086
.235
t=300 11/20/75
20.5
9.1
.089
.235
Precipitated	Calculated	(l.Og Seed)
	(SO	Solids	Solids	Rel. Saturation
.250	.115 				1-864
.250	.093 8.58g	12.04g	1.441
.250	.081 11.7 5g	16.63g	1.245
.250	.079 11.73g	20.64g	t. 203
.250	.077 12.59g	21.79g	1.164
.250	.074 12.66g	20.64g	i.^g
.250	.072 12.67g	21.21g	i.ng
.250	.070 12.47g	21.79g	1.087
.250	.0700 12.llg	20.07g	i.093
.250	.067 8.44g			1.078

-------
TABLE 3-1 (Cont'd)
RUN NO. _3	
MgSO3-6H20 @ 30°C
(Min.)
Tiw/Date
Te*p °C
PH
(MR++)
(Na+)
t=0 10/15/75
32.5
9.04
.147
.232
t-31 10/15/75
32.8
9.00
.135
.232
t-82 10/15/75
32.5
8.97
.119
.232
t=130 10/15/75
32.1
8.98
.115
.232
t-191 10/15/75
32.1
8.90
.115
.232
t»251 10/15/75
32.1
9.00
.115
.232
Precipitated Calculated	^8 Seed)
(CI")	(SO, = )	Solids	Solids Rel. Saturation
. 294	.116	— 		1.434
.294	.107	5.77g	7.24g	1.283
.294	.091	22.18g	18.43g	1077
.294	.089	20.44g	21.10g	1.055
.294	.089	15.48g	21.10g	1.055
.294	.089	22.75g	21.10g	1055

-------
TABLE 3-1 (Cont'd)
RUN NO. A
MgSOi¦6H2O 0 30°C
(Min.)
Tine/Date
Tesp °C
PH
(MR++)
(Na+)
t=0 10/21/75
32.7
9.00
.109
.216
t=40 10/21/75
32.7
9.02
. 107
.216
t=120 10/21/75
31.9
9.05
.1045
.216
t=215 10/21/75
31.5
9.10
.102
.216
t»275 10/21/75
31.5
9.05
.100
.216
t=334 10/21/75
31.5
9.10
.100
.216
VO
ON
Precipitated Calculated ^6 Seed)
(Cl~)	(SO; *)	Solids	Solids	Rel. Saturation
218
.108
—
	
1.179
218
.105
1.35g
1 ¦ 32g
1.145
218
.1035
4.31g
2. 96g
1.145
218
.101
4.77g
4.61g
1.126
218
.0995
5.69g
5.60g
1.107
218
.100
6. 95g
5.92g
1.108

-------
TABLE 3-1 (Cont'd)
RUN NO. 5
MgSOj- 6H20 9 30°C
(Min.)
Tiae/Date	Te»p °C	gH	(He**)	(Ka4-)	
vO
t=0 10/30/75	31.1	9.15	.164	.307
t=25 10/30/75	30.8	9.15	.113	.307
t=40 10/30/75	30.8	9.12	.1065	.307
t=60 10/30/75	31.0	9.10	.1065	.307
t-120 10/30/75	30.9	9.15	.107	.307
t=240 10/30/75	30.9	9.15	.1045	.307
t«280 10/30/75	30.6	9.16	.1045	.307
t"=5700 11/3/75	30.6	9.12	.1085	.307
Precipitated Calculated (1.02g Seed)
(CI )	(SO, )	Solid3	Solids	Rel. Saturation
.329	.1529	--	—	1.858
.329	.098	38.86g	J3.90g	1.109
.329	.0975	38.OOg	38.50g	1.067
.329	.096	39.39g	38.50g	1.049
.329	.0965	38.41g	38.50g	1.059
.329	. 096	38.43g	39.80g	1.040
.329	.091	36.50g	39.80g	1.011
.329	.093	42.89(5	—	1.050

-------
TABLE 3-1 (Cont'd)
RUN NO. 6
MgS03-6H20 £ 40°C
(Mln.)
Tine/Date
Te«p °C
PH

-------
TABLE 3-1 (Cont'd)
RUN NO. 7
MgSOJ • 6H2O 9 40°C

-------
TABLE 3-1 (Cont'd)
KUN NO. 8
MgS02'3Hj0 « 55°C
(Mln.)
Time/Date	leap °C	gH	(Mn++)	(«a+)
IO
O
O
t-0 10/22/75	55	8.23	.1427	.2754
t-680 10/23/75	55	8.23	.142	.2754
t=2880 10/24/75	55	8.23	.142	.2754
t=4320 10/25/75	55	8.15	.141	.2754
t=5700 10/26/75	46	8.00	.141	.2754
t-7200 10/27/75	52	7.98	.141	.2754
Precipitated Calculated (.l25g Seed)
(Cl~)	(S03~)	Solids	Solids Rel. Saturation
.2854	.1377	—	1-550
.2854	.135	1.06g	0.34g	1.526
.2854	.135	1.30g
.2854	.136	2.45g
.2854
.2854
0-698	1.522
0848	1.526
.133	2.49?	0-Mg	1.358
.132	3.15g	°-84g	1.438

-------
TABLE 3-1 (Cont'd)
RUN NO. 9
MgS03-3H20 055°C
(Mill.)
Tine/Date	Tewp °C	gH	(Hg++)	(Wa+)	
t-0 11/3/75	54.9	8.18	.125	.244
t=1440 11/4/75	54.9	8.18	.125	.244
t-2880 11/5/75	53.9	8.18	.125	.244
t-4320 11/6/75	55.5	8.05	.130	.244
t=5700 11/7/75	55	8.01	.130	.244
t=11520 11/10/75 54.5	7.85	.123	.244
N5
O
.	(Seed Solution =
Precipitated Calculated	i.0362b/500b1)
(CI )	(SOj~)	Solids	Solids	Rel. Saturation
.250 .122 -- ""	1-351
.250 .118 1.39g ""	1.Ml
.250 .119 1.61g "	I-313
.250 .119 1.70g ""	1-366
.250 .119 1.35g "	1-357
.250 .111 1.94g "	1-243

-------
TABLE 3-1 (Cont'd)
RUN NO. 10
MgS03-3H20 0 70°C
(Min.)
Time/Date
Teap °C
PH
(Mr++)
(Na+)
t=0 11/17/75
70.5
7.50
.172
.340
t=10 11/17/75
71
7.30
.170
.340
t=20 11/17/75
71
7.1
.165
.340
t=30 11/17/75
71
7.0
.163
.340
t=40 11/17/75
71
7.2
.160
.340
t«50 11/17/75
71
7.0
.157
.340
t=60 11/17/75
71
7.0
.157
.340
t»120 11/17/75
71
6.9
.151
.340
t=240 11/17/75
70.5
6.8
.140
.340
t-1680 11/18/75
70.1
6.75
.132
.340
t-3120 11/19/75
71
6.60
.139
.340
t=4560 11/20/75
71
6.90
.145
.340
t=6000 11/21/75
71
6.80
.155
.340
N)
O
K>
(Seed Solution =
Precipitated Calculated 5.42g/500i»l)
(Cl~)	(SO, **)	Solids	Solids	Rel. Saturation
.344	.170 				2.196
.344	.162	5.99g lllg	2.087
.344	.158	7.63g 3.88g	1.986
.344	.158	8.23g 4-99B	1.944
.344	.149	9.40g 6-65B	1.898
.344	.150	4.84g 8.32g	1.837
.344	.148	4.78g 8,32S	1.821
.344	.108	11.22g	11.64g	x 395
.344	.134	10.61g	17.74g	t 550
.344	.145	14.85g	22.17g	t 297
.344	.102	6.21g	18.29g	1-179
.344	.096	9.61g 		1.247
.344	.0886	10.52g ™	1.180

-------
TABLE 3-1 (Cont'd)
RUN HO. 11
HgSO,-3H20 0 85°C
(Hin.)
Tine/Date
Teap °C
PH
(M8++)
(Na+)
t=0 12/15/75
86
8.1
.124
.240
t-10 12/15/75
86
8.0
.124
.240
t=20 12/15/75
85
8.0
.117
.240
t-30 12/15/75
84
7.8
.108
.240
t«40 12/15/75
84
7.6
.100
.240
t-50 12/15/75
83
7.5
.097
.240
t=60 12/15/75
84
7.5
.098
.240
t-120 12/15/75
84
7.3
.094
.240
. t=180 12/15/75
N
O
85
7.3
.091
.240


•

t*> t=1620 12/16/75
85
7.3
.081
.240
(CI")
(S
-------
TECHNICAL NOTE 200-045-54-01
TRANSITION OF MgS03 HYDRATES
IN 3M MgClz SOLUTION
Prepared by:
J. L. Skloss
204

-------
FIGURES
Number	page
1	MgSCh Solids Present in the 3M MgCl2 Solution,
200X	 207
2	DSC Scan of the MgSOs Solids Present in the 3M
MgCl2 Solution	 208
205

-------
It has been found theoretically that the transition temper-
ature of MgS03*3H20 and MgS03*61120 can be lowered to room
temperature in a medium of 3M magnesium chloride solution. The
trihydrate phase of magnesium sulfite is predicted to be stable
at 25°C in this medium. This prediction was verified by ex-
periment .
A 1-liter solution of 3M magnesium chloride was prepared,
and excess MgS03-61120 solids were added to saturate the solution
with respect to magnesium sulfite at room temperature. After
filtration, lOg of MgS03*6H20 and lOg of MgS03*3H20 solids were
added. The solids were allowed to equilibrate with the 3M
magnesium chloride solution at room temperature (25±2°C) with
gentle stirring on a magnetic stirrer. A stoppered 1-liter
Erlenmeyer flask was used to contain the system, and a nitrogen
purge was used as necessary to prevent oxidation of sulfite ion.
After eight days of continuous stirring, a 50-ml portion
of the slurry was filtered on Whatman No. 42 paper. The solids
were washed with 507o ethanol and over-dried at 45°C for 30 min-
utes. The chemical analyses of the solids showed that the
magnesium and the sulfite concentrations were 6.31±0.05 mmole/g,
which was expected for MgS03*31120. The trihydrate phase of
magnesium sulfite was identified in the microphotograph (see
Figure 1). Also the DSC scan, Figure 2, indicated that the
solids sample was MgS03«3H20.
206

-------


•< 1
t
. \
. «r-*
f £
• i
- •
¦ *
•

• V*

FIGURE 1 MgS03 SOLIDS PRESENT IN THE
3M MgCl2 SOLUTION, 200X
207

-------
nun wo. wooti
hum No.ioATf'tLy [TAXIS
ERA TOR		(SCALE. tVtn	*>
8AMPLF: "«•»-)**-V«^|P00G BSTt -CJr^	
_ _	IHEAT	COOL	«SO_
ATM T& *•	I SHIFT, kl	*S_
FLOW RATE	
DTA-DSC
SCALE. "C/ln I 0 _ .
I mCdl/sBC t/fr)		
weoHT. >"o ' 9f
Wfct-fcMfcNCE	j
TGA
3CALF. rnp/in 		
S(_W*FT»frSSfON, mp_
WEIGHT, 	— -
TtMB COH&T . *«_ _
dV. C rrtg/mlr* Ifm 		
TMA
SCALE. m»n/ir,	
MODE		
SAK*cn_e size	
LOAD, g	
cfY. C10X I. (mUs/rr^n V»r» _
:/0^	
'tt" 	aB""",ao wi	sd
TEMPERATUnP, XT ( CHWOMI-t 7ALUMEL J
Figure 2. DSC Scan of the MgS03 Solids Present in the 3M MgCl2 Solution

-------
TECHNICAL NOTE 200-045-54-02
MgSO 3 HEXAHYDRATE AND TRIHYDRATE
PREPARATION, HANDLING
AND CHARACTERIZATION
Prepared by:
R. E. Sawyer
J. L. Skloss
R. E. Pyle
209

-------
CONTENTS
1.	Introduction	212
2.	Preparation of Magnesium Sulfite
Hexahydrate and Trihydrate 	 213
Nucleation of MgSO3 Hexahydrate 	 213
Nucleation of MgSO 3 Trihydrate	213
Conversion of Hydrates 	 214
3.	Solids Handling	215
4.	Solids Characterization 	 217
DSC Theory of Operation and
Application	217
Theory and Application of TGA
Analysis	223
Dessicator TGA	226
Bibliography 	 231
Number	FIGURES	Page
4-1 DSC Scan of MgS03-3H20	219
4-2 DSC Scan of MgS03*6H20 		220
4-3 DSC Scan of MgS03«6H20 and MgS03*3H20 		222
4-4 TGA of MgS03 • 6H20 and MgS03«3H20 Mixture	224
4-5 MgS03*6H20 Wt. Loss vs. 7a Composition @ 100°C
(self generated atmosphere)	229
210

-------
Number	TABLES	Page
4-1 	 221
4-2 Dessiccator TGA Results (§ 100°C	 228
211

-------
SECTION 1
INTRODUCTION
The removal of S02 from flue gas streams by an MgO regenera-
tive scrubbing system has proven to be technically feasible.
The S02 is removed from the scrubbing system as a precipitate of
hydrated magnesium sulfite. Two hydrates have been identified,
MgO3•6H20 and MgS03-3H20.
In order to make experimental measurements in support
of the magnesium oxide process, it is necessary to have sample
taking, handling and preparation methods such that the material
actually analyzed is representative of the process from which
it was sampled. This technical note describes such procedures
developed for the tri- and hexahydrates of MgS03.
212

-------
SECTION 2
PREPARATION OF MAGNESIUM SULFITE HEXAHYDRATE AND TRIHYDRATE
Pure crystals of magnesium sulfite hexahydrate or trihydrate
may be prepared by two methods: nucleation and hydrate con-
version. Either hydrate of MgS03 may be prepared by nucleation,
depending on temperature, from supersaturated solutions of mag-
nesium sulfite. Once a particular hydrate phase is obtained,
conversion to the other hydrate can be achieved by temperature
adjustment and allowance for equilibration.
NUCLEATION OF MgS03 HEXAHYDRATE
Pure MgS03*6H20 may be nucleated from magnesium sulfate
solution by the addition of concentrated sodium sulfite solu-
tion. A solution temperature of 30°C is suitable. When suffi-
cient sodium sulfite has been added (^0.5m) nucleation occurs.
After the onset of nucleation approximately 30 minutes is allowed
for the precipitation of solids. 10-100 micrometer sized
MgS03*6H20 crystals are obtained by this method.
NUCLEATION OF MgS03 TRIHYDRATE
Pure MgS03'3H20 may be nucleated by the addition of con-
centrated sodium sulfite solution to magnesium sulfate solution
heated above 65°C. Extremely small (1-5 micrometer) crystals
are obtained by homogeneous nucleation under these conditions.
Larger crystals up to 50 micrometers may be obtained by over-
night stirring of the slurry and subsequent precipitation of
the solid.
213

-------
CONVERSION OF HYDRATES
Magnesium sulfite hexahydrate may be converted to the
trihydrate phase and vice versa by adjusting the temperature
of the slurry containing an excess of magnesium sulfite solids.
MgS03*6H20 is stable in pure water slurries below A1°C whereas
MgS03*3H20 is stable above 41°C. Increasing the ionic strength
of the solution decreases the transition temperature. The
hexahydrate phase is converted completely to the trihydrate
phase overnight by continually stirring the slurry maintained
at approximately 55°C. This conversion takes place either in
pure solutions of magnesium sulfite or in 20"L magnesium sulfate
medium. The trihydrate phase of magnesium sulfite is converted
to the hexahydrate phase by overnight stirring at room tempera-
tures .
214

-------
SECTION 3
SOLIDS HANDLING
Once a pure hydrate form of magnesium sulfite is prepared
in a slurry medium it is important to remove the solid phase
from the solution without affecting the integrity of the cry-
stals. The most applicable method is vacuum filtration followed
by alcohol washings. Only a few seconds are required to filter
100 ml of slurry through a 47-ram filter membrane. Trihydrate
slurries filter more slowly than hexahydrate slurries because
of the smaller crystal size. The filter membranes may consist
of glass fiber, paper, or other common filtering materials
which are resistant to alcohol.
After all of the slurry sample has been filtered, the mag-
nesium sulfite crystals are washed immediately with 50 vol.7o
methanol or ethanol in water solutions that have been preheated
to the slurry temperature. Three 10-ml portions of alcohol
solution are recommended for the complete washing of the cry-
stals. Tests with barium chloride addition showed that all
the sulfate ion from the 20% magnesium sulfate medium is removed
from the crystals by these washings.
The 50% alcohol solution is a very suitable washing sol-
vent. The solubility of magnesium sulfite in 50% ethanol is
only 1/30 of that in pure water at the temperatures encountered
in this study. Also, the removal of magnesium sulfate, sulfite
or chloride solution from the magnesium sulfite crystals is
easily and quickly accomplished by the alcohol solution washings.
After washing, the crystals present on the filter membrane
are placed in an oven set at 40-50°C. Solid samples less than
1 gm dry within 30 minutes. More time is required for larger
215

-------
samples, and overnight drying may be necessary in some instances.
Caution should be used not to exceed 50°C; otherwise, some of
the waters of hydration may be lost. The dried solids are gen-
erally stored in a desiccator (without desiccant) purged with
nitrogen or other inert gas. The magnesium sulfite solids may
also be stored in air without problem provided the relative
humidity does not exceed 90%. No crystal deterioration of either
hydrate phase was observed during four months storage time in
plastic bottles.
216

-------
SECTION 4
SOLIDS CHARACTERIZATION
The Radian Literature Survey Technical Note 200-045-36-01
reported that there have been several methods used to charac-
terize magnesium sulfite hexahydrate and trihydrate solids.
Such characterization techniques as infrared (IR), X-ray, dif-
ferential thermal analysis (DTA) and differential scanning
colorimetry (DSC) have been used successfully (D0-015, DA-195,
TE-030). Of the analytical methods used, DSC appears to offer
the greatest ease of operation coupled with a rapid turn-around
time. A Dupont 990 thermal analysis system consisting of a 990
programming and recording console and a 990 system cell base
with standard DSC cell were used in our laboratory.
DSC THEORY OF OPERATION AND APPLICATION
In the DSC mode, the Dupont System 990 gave very accurate
results. Two identical aluminum pans, one containing the sam-
ple to be analyzed, the other the reference, are placed on the
DSC cell. The cell is then covered and sealed to form an in-
sulated environment. A programmed heating rate and maximum
temperature limit are then selected on the 990 program and re-
cording console. The system is then operator balanced to adjust
the electronics to the temperature indicated inside the sample
chamber by the thermocouples. The system is then placed in
operation and allowed to heat at its programmed rate. As the
sample and reference pans heat, the rate of energy absorption
or evolution by the sample is recorded on a dual pen recorder
as a function of temperature vs. change. To perform kinetic
studies, the displacement of the curve from some predetermined
or predrawn baseline is measured.
217

-------
In the DSC thermogram of MgS03*3H20, shown in Figure 4-1,
only one endothermic transition was observed, starting at 120°C
with a peak at 155°C. In the DSC scan of MgS03*61120, shown in
Figure 4-2, it is significant to note that only one endothermic
transition was observed starting at 45°C with a peak maximum
near 90°C. These results are consistent with the studies of
Dauerman (DA-195).
In a further refinement of the DSC studies, the area of
the peaks generated for both pure MgS03*6H20 and MgS03*3H20 of
various weight ranges were determined. From this a correlation
factor was empirically determined. This correlation factor
times the area of the peaks gave a very good approximation
(within a few hundredths of a mg) of the actual amount in mg
of each hydrate present.
The correlation factors for both MgS03*6H20 and MgS03*3H20
were determined by an area approximation technique of curves
generated by pure samples. DSC scans were made of pure samples
of either MgS03*6H20 or MgS03*3H20 which had been weighed on
an analytical balance. The area of the peak generated by each
species was calculated by multiplying the peak height in inches
by the peak width at one-half the peak height. This was then
divided into the total sample weight to determine the factor
in mg/in2. Typical results are shown in Table 4-1.
This approach was tried with synthetic mixtures of pure
MgS03«6H20 and MgS03*3H20. Typical scans such as Figure 4-3
were produced for these mixtures. By applying the aforemen-
tioned techniques, it is possible to determine the amounts of
hexa- and tri-crystals in a mixture to within a total accuracy
of a few percent.
218

-------
Run No: 6
Date: 3-17-76
Operator: RES
Sample: P. S. Sample
Composite Sample Dried Solid
1/7/76
DTA-DSC
Scale, °C/in: 10
Ueight, mg: .00536 gr
Reference: Empty Pan
* - ¦ * 1 ¦ '	« i i t i i «	1	•	1	1 . i
20 40 60 80 100 120 140 160 180 200
Temperature, °C
Figure 4-1. DSC Scan of MgS03*3H20
02-2300-1

-------
Run No: 1
Date: 3-22-76
Operator: RES
Sample: MgS03*6H20
DTA-DSC
Weight, mg: .8
Reference: Empty Pan
i i	« i i i	 			i	•	i	 ¦ ¦ ¦
20 40 60 80 100 120 140 160 180 200
Temperature, °C
Figure 4-2. DSC Scan of MgS03'6H20

-------
TABLE	4-1.
Sample
Weight	i Peak Area
Species	in mg.	in sq. in.	mg/in2
MgSO 3•6H2O .80	3.6	.22
MgSO 3•6H20 2.36	9.13	.26
HgS03•6H20 .46	1.8	.26
MgSO3 • 6H2O .58	2.3		._25_
Average	.25
MgSO 3•3H20 .85	2.4	.35
MgSO 3•3H20 .63	2.1	.30
MgSO 3•3H20 .63	1.9	.33
MgSO 3•3H20 1.11	3.2	.34
Average	'33
221

-------
Run No: 4
Date: 3-18-76
Sample: MgS03•6H2O
MgS03'3H20
Atm: Closed Pan
DTA-DSC
Scale, °C/in: 10
Weight, Mg: 0.98
Reference: Empty Pan
	« 1	¦ 1	1 1 «	1			« . » .	
20 40 60 80 100 120 140 160 180 200
Temperature, °C
Figure 4-3. DSC Scan of MgS03#6H20 and MgS03*3H20

-------
THEORY AND APPLICATION OF TGA ANALYSIS
Another analytical technique which is useful in quantifying
mixtures of the tri- and hexahydrates is thermogravimetric anal-
ysis. A Dupont 990 thermal analyzer control and recording con-
sole was used in conjunction with a Dupont 951 thermogravimetric
analyzer (TGA). The heart of the TGA system consists of a pre-
cision microbalance and counterweighting system which can be
insulated from the surrounding environment. A sample of up to
1 gram is placed on the balance pan and is coarsely counterbal-
anced on the balance arm by. applying counterweights. The entire
system is then isolated from the surrounding environment. A
"fine tune" balance is then accomplished by means of electronic
counterbalancing. A constant heating rate is then applied to
the sample in the balance pan. The 990 control unit will
record weight loss or gain as a function of temperature.
The accuracy of the method was investigated by analyzing
the thermograms of synthetic mixtures of the hexahydrate and
tri-hydrate phases ranging from 10-90% in composition. From
these thermograms, the water content and the magnesium sulfite
content can be calculated. First, the water content calcula-
tions will be considered and compared to the theoretical values.
The water content of each hydrate in a mixture can be
obtained from the TGA thermogram (a typical thermogram is shown
in Figure 4-4) by taking into account the fact that the weight
loss in the first dehydration step at 100°C represents the first
three moles of water in MgS03*6H20, i.e., 50% of the water con-
tent of the hexa form. The water loss beginning at 180°C is the
remainder of the hexahydrate water plus all of the trihydrate
water.
223

-------
\MPLE: Std #3
18.96% MgS03*3H20
81.04% MgS03*6H20
'lg. No. 11
ZE..15 mg.
X-AXIS
Y-AXIS
RUN NO. _ 1	Da» ..
OPERATOR . RM	
HEATING RATE. 10 .*C._
mk».
ATM. self-generated
TIME CONSTANT 1 . »cc. .
TEMP. SCA1.E SO *C
inch
SHIFT 9 inch
TIME SCALE (ALT.l
SCALE .'2 . _ /"9_
inch
(scalc Getting x 2)
SUPPRESSION ... 50 mg.


























































































































1



























































































































Temperature* °C
Figure 4-4. TGA of MgG03c6H20 and MgS03 *31120 Mixture

-------
The amount of each hydrate may be calculated as follows:
7» H20 in MgS03«6H20 in a mixture = (7» weight loss in the
first step to 175°C) x (2)
% H20 in MgS03*3H20 in a mixture = (7* weight loss in the
second step to 400°C) - (% weight loss in the first
step to 175°C)
The theoretical values of the water content are calculated
as follows:
% H20 in MgS03*6H20 in a mixture - % MgS03-6H20 in the mixture
v 6H,0
x MgS03•6H20
-	% MgS03«6H20 in the mixture
v 108
x nn
-	MgS03*6H20 in the mixture
x .509
7o H20 in MgS03*3H20 in a mixture ¦ 7o MgS03*3H20 in the mixture
3HjO
x MgS5j.3H,6
% MgS0j*3H20 in che mixture
x 158.3
225

-------
= 7<> MgS03*3H20 in the mixture
x 0.341
The magnesium sulfite content of a sample can be calculated
from the TGA thermograms by performing a similar procedure as
the water content calculations, and by taking into account that
6 moles of water represent 50.97o by weight of MgS03*6H20 and 3
moles of water represent 34.1 7o by weight of MgS03»3H20. Thus
the percentage of each hydrate in a mixture is calculated as
follows:
7o MgS03*6H20 in a mixture = % weight loss at 175°C x
2/0.509
7» MgS03*3H20 in a mixture = (7. weight loss between 175°
and 400°C - 7o weight loss at 175°C)/34.1
DESSICATOR TGA
While both DSC and TGA offer excellent analytical means
for determining the compositions of mixtures of hexahydrate and
trihydrate, neither method is well suited for field determina-
tions. The DSC system is moderately expensive (-$10,000) and
not well suited for analyzing a large number of samples since
it takes approximately 30 minutes for a single analysis. The
TGA analysis while slightly more accurate is also more prone
to operator error. The micro-balance system would also not
adapt easily to a field environment, and the time required for
a single analysis is longer than with DSC. Therefore, a modi-
fied field TGA method was developed.
226

-------
A large glass vacuum desiccator was fitted with a heating
mantle controlled by a Variac. All desiccant was removed and
a thermometer was placed on the dessicator plate. The Variac
setting was adjusted until the thermometer inside registered
between 110-110°C. According to Dauerman (DA-195), when
MgS03*6H20 is heated in a self-generated atmosphere, two transi-
tion steps are observed: MgS03*6H20 MgS03 *31120 starting at
100°C, then MgS03*3H20 -* MgS03 anhydrous starting at 175°C.
Therefore, by adjusting the inside desiccator temperature to
100-110°C, the hexahydrate form should convert to trihydrate,
but the trihydrate should remain unchanged.
A number of oven-dried glass vials were labeled and weighed
on an analytical balance. To each vial, approximately 0.6 grams
of pure hexahydrate or trihydrate was added. The exact weight
added was determined by difference on the analytical balance.
A glass wool plug was then inserted into each vial to maintain
a self-generated atmosphere. The samples were then reweighed
and placed into the heated glass desiccator. After 4 hours,
all the samples were removed and cooled in an argon purged
desiccator. The samples were then reweighed. The average
weight loss was approximately 25% for hexahydrate samples and
less than 1% for trihydrate samples (Table 4-2). The theoreti-
cal weight loss for the hexahydrate is 25.4%. Results for both
tests are shown in Figure 4-5.
There are a number of advantages to the desiccator TGA
method. A large number of samples can be processed at the same
time, since approximately 20 samples can be placed in the desic-
cator. The weighing time for each sample is less than 10 min-
utes. Only an analytical balance and several desiccators are
required for the analysis. A disadvantage, however, is that
the technique is insensitive to contamination from inert species;
that is, material that cannot be removed from the MgS03 hydrate
227

-------
TABLE 4-2. DESSICCATOR TGA RESULTS § 100°C
Sample, wt% Grams	Wt. Loss, %
Hexa	Tri	Sample Wt.	Wt. Loss	Observed	Theoretical
100	0	0.599	0.152	25.4	25.4
100	0	0.602	0.153	25.4	25.4
0	100	0.602	0.0018	0.29	0.0
0	100	0.600	0.0005	0.08	0.0
60	40	0.600	0.093	15.4	15.2
60	40	0.603	0.092	15.3	15.2
80	20	0.602	0.120	19.8	20.3
80	20	0.535	0.105	17.5	20.3

-------
1007.
807. MgSOj • 6HaO
207. MgSOi.3HaO
« 15
90 100
7. MgSOfSHzO in sample
Figure 4-5. MgS03*6H20 Wt. Loss vs. 7. Composition @ 100°C
(self generated atmosphere)
229

-------
sample by such techniques as alcohol/water washing. An accurate
estimate of the amount of contamination can be determined by
wet chemical analyses.
230

-------
BIBLIOGRAPHY
DA-195 Dauerman, Leonard, Sulfur Oxide Removal From Power
Plant Stack Gas. Magnesia Scrubbing-Regeneration:
Production of Sulfuric Acid. Newark, New Jersey, New
Jersey Institute of Technology, Feb. 1975.
DO-015 Downs, W. and A. J. Kubasco, Magnesia Base Wet
Scrubbing of Pulverized Coal Generated Flue Gas --
Pilot Demonstration. PB 198 074 Alliance, Ohio,
Babcock and Wilcox Co., 1979.
TE-030 Tennessee Valley Authority, Removal of Sulfur Dioxide
from Stack Gases. Thermal Decomposition of Magnesium
Sulfite. Muscle Shoals, Ala., Feb. 1971, #38.
231

-------
TECHNICAL NOTE 200-045-54-03a
PRECIPITATION KINETICS OF MgS03
HYDRATES IN SCRUBBER-LIKE MEDIA
Prepared by:
J. L. Skloss
T. B. Parsons
P. S. Lowell
232

-------
CONTENTS
1.	Introduction	240
2.	Experimental Approach 		241
3.	Data Analysis Methods	243
4.	Results and Discussion	248
Experiment 3-2, Trihydrate Seed
Crystals	261
Experiment 3-3, Hexahydrate Seed
Crystals	263
Experiment 3-4, Trihydrate Seed
Crystals	263
Experiment 3-5, Hexahydrate Seed
Crystals	266
Experiment 3-6, Mixed Tri- and Hexa-
hydrate Seed Crystals	268
Experiment 3-7, Mixed Tri- and Hexa-
hydrate Seed Crystals	268
5.	Conclusions	271
Appendix A	272
233

-------
FIGURES
Number	Page
4-1 Results of DSC Analysis of a Pure Sample of
MgS03*6H20 	 255
4-2 Results of DSC Analysis of a Pure Sample of
MgSO 3 • 3H20 	 256
4-3 Results of DSC Analysis of Solids Precipitated
After 470 Minutes in Experiment No. 3-5	 257
4-4 Electron Micrograph of Pure MgS03*6H20 at 300X
Magnification	 258
4-5 Electron Micrograph of Pure MgS03*3H20 at 1000X
Magnification	 259
4-6 Electron Micrograph of Solids from Experiment
No. 3-5 at 1000X Magnification Showing
Trihydrate Crystals	 260
4-7 Results of MgS03 Hydrate Precipitation Rate
Experiment 3-2 Employing Trihydrate Seed
Crystals	 262
4-8 Results of MgS03 Hydrate Precipitation Rate
Experiment 3-3 Employing Hexahydrate Seed
Crystals	 264
4-9 Results of MgS(>3 Hydrate Precipitation Rate
Experiment 3-4 Employing Trihydrate Seed
Crystals	 265
234

-------
FIGURES (Continued)
Number	Page
4-10 Results of MgS03 Hydrate Precipitation Rate
Experiment 3-5 Employing Hexahydrate Seed
Crystals	 267
4-11 Results of MgS03 Hydrate Precipitation Rate
Experiment 3-6 Employing a Mixture of Tri- and
Hexahydrate Seed Crystals	 269
4-12 Results of MgS03 Hydrate Precipitation Rate
Experiment 3-7 Employing a Mixture of Tri- and
Hexahydrate Seed Crystals	 270
A-l Results of DSC Analysis of Solid Sample Taken
at 60 Minutes in Experiment 3-2	 273
A-2 Results of DSC Analysis of Solid Sample Taken
at 3150 Minutes in Experiment 3-2	 274
A-3 Results of DSC Analysis of Solid Sample Taken
at 10 Minutes in Experiment 3-3	 275
A-4 Results of DSC Analysis of Solid Sample Taken
at 20 Minutes in Experiment 3-3	276
A-5 Results of DSC Analysis of Solid Sample Taken
at 60 Minutes in Experiment 3-3	 277
A-6 Results of DSC Analysis of Solid Sample Taken
at 150 Minutes in Experiment 3-3	 278
235

-------
FIGURES (Continued)
Number	Page
A-7 Results of DSC Analysis of Solid Sample Taken
at 200 Minutes in Experiment 3-3	279
A-8 Results of DSC Analysis of Solid Sample Taken
at 60 Minutes in Experiment 3-4	280
A-9 Results of DSC Analysis of Solid Sample Taken
at 1260 Minutes in Experiment 3-4	281
A-10 Results of DSC Analysis of Solid Sample Taken
at 20 Minutes in Experiment 3-5	282
A-ll Results of DSC Analysis of Solid Sample Taken
at 60 Minutes in Experiment 3-5	283
A-12 Results of DSC Analysis of Solid Sample Taken
at 100 Minutes in Experiment 3-5	284
A-13 Results of DSC Analysis of Solid Sample Taken
at 120 Minutes in Experiment 3-5	285
A-14 Results of DSC Analysis of Solid Sample Taken
at 150 Minutes in Experiment 3-5	286
A-15 Results of DSC Analysis of Solid Sample Taken
at 180 Minutes in Experiment 3-5	287
A-16 Results of DSC Analysis of Solid Sample Taken
at 270 Minutes in Experiment 3-5	288
236

-------
FIGURES (Continued)
Number	Page
A-17 Results of DSC Analysis of Solid Sample Taken
at 330 Minutes in Experiment 3-5	289
A-18 Results of DSC Analysis of Solid Sample Taken
at 390 Minutes in Experiment 3-5	290
A-19 Results of DSC Analysis of Solid Sample Taken
at 470 Minutes in Experiment 3-5	291
A-20 Results of DSC Analysis of Solid Sample Taken
at 1400 Minutes in Experiment 3-5	292
A-21 Results of DSC Analysis of Solid Sample Taken
at 3200 Minutes in Experiment 3-5	293
A-22 Results of DSC Analysis of Solid Sample Taken
at 20 Minutes in Experiment 3-6	294
A-23 Results of DSC Analysis of Solid Sample Taken
at 50 Minutes in Experiment 3-6	295
A-24 Results of DSC Analysis of Solid Sample Taken
at 60 Minutes in Experiment 3-6	296
A-25 Results of DSC Analysis of Solid Sample Taken
at 180 Minutes in Experiment 3-6	297
A-26 Results of DSC Analysis of Solid Sample Taken
at 300 Minutes in Experiment 3-6	298
237

-------
FIGURE (Continued)
Number	Page
A-27 Results of DSC Analysis of Solid Sample Taken
at 390 Minutes in Experiment 3-6		299
A-28 Results of DSC Analysis of Solid Sample Taken
at 4260 Minutes in Experiment 3-6		300
A-29 Results of DSC Analysis of Solid Sample Taken
at 10 Minutes in Experiment 3-7		30^
A-30 Results of DSC Analysis of Solid Sample Taken
at 30 Minutes in Experiment 3-7		302
A-31 Results of DSC Analysis of Solid Sample Taken
at 40 Minutes in Experiment 3-7	 303
A-32 Results of DSC Analysis of Solid Sample Taken
at 80 Minutes in Experiment 3-7		304
A-33 Results of DSC Analysis of Solid Sample Taken
at 150 Minutes in Experiment 3-7		305
A-34 Results of DSC Analysis of Solid Sample Taken
at 240 Minutes in Experiment 3-7	 306
A-35 Results of DSC Analysis of Solid Sample Taken
at 320 Minutes in Experiment 3-7	 307
A-36 Results of DSC Analysis of Solid Sample Taken
at 1400 Minutes in Experiment 3-7	 30g
238

-------
TABLES
Number	page
2-1 Summary of Experimental Conditions in Tests of
Magnesium Sulfite Hydrate Precipitation from
Scrubber-Like Media	 242
4-1 Results of Magnesium Sulfite Hydrate Precipita-
tion Experiment No. 3-2	 249
4-2 Results of Magnesium Sulfite Hydrate Precipita-
tion Experiment No. 3-3	 250
4-3 Results of Magnesium Sulfite Hydrate Precipita-
tion Experiment No. 3-4	 251
4-4 Results of Magnesium Sulfite Hydrate Precipita-
tion Experiment No. 3-5	 252
4-5 Results of Magnesium Sulfite Hydrate Precipita-
tion Experiment No. 3-6	 253
4-6 Results of Magnesium Sulfite Hydrate Precipita-
tion Experiment No. 3-7	 254
239

-------
SECTION 1
INTRODUCTION
The work described in this technical note is part of a
study of the physical chemistry of the precipitation of mag-
nesium sulfite hydrates. This chemical phenomenon is important
in the design of magnesium-based SO2 wet scrubbing processes.
Field tests of the magnesium oxide processes have shown that
both MgS03 trihydrate and hexahydrate precipitate from scrubbing
solutions. The laboratory tests described here were done to in-
vestigate the rates of precipitation of the two hydrates from
slurries characteristic of those in the magnesium oxide process.
Precipitation rate studies in dilute solutions and experiments
to study nucleation rates have also been performed and are
described in Technical Notes 200-045-36-03, -54-04, 05 and 06.
240

-------
SECTION 2
EXPERIMENTAL APPROACH
The equipment and experimental procedure employed in the
precipitation rate studies are described in Technical Note
200-045-36-05, which gives the results of identical precipita-
tion rate experiments in dilute solutions. A 4-liter pyrex
reaction vessel was employed. The well-stirred reactor was
situated in a constant temperature bath. All tests were done
at 55°C. Batch addition of MgSOi* and Na2S03 feedstock solutions
and MgS03 seed crystals was employed. The reactant solutions
were composed of approximately 20 wt percent (2.08 molal) MgSO^
and were supersaturated with respect to magnesium sulfite.
The initial sulfite concentration was varied from about 0.3
to 0.5 gmole/2, by varying of the concentration of the Na2S03
feed solution. Precipitation was initiated by the addition of
1 gram of MgS03*6H20 or MgS03*3H20 seed crystals. In some ex-
periments a mixture of 0.5 grams of each hydrate was employed.
The experimental conditions of 55°C, pH 6, and 20 wt % MgSOt*
solutions are characteristic of solutions in operating magnesium
oxide processes (scrubber-like media).
Reactor contents were sampled before addition of seed
crystals and periodically thereafter. Tests were conducted for
periods of 1,100 to 11,000 minutes. Analyses were performed to
measure sodium, magnesium, total sulfur, sulfite, and sulfate
(by difference) concentrations in the liquor; weight percent
solids in the slurry; pH and temperature. Sulfate determina-
tions indicated that the amount of sulfite oxidation to sulfate
was negligible. The wet analytical chemistry methods listed in
section 2.3 of Technical Note 200-045-36-05 were employed. In
addition, microscopic examination and differential scanning
241

-------
calorimetric (DSC) analyses of solids were performed. These
methods are described in Technical Note 200-045-54-02.
It was found that measurement of the weight of solids
produced was not a reliable indicator of the reaction rate.
The results of chemical analyses of sulfite in the solution
were used to calculate the number of moles of MgS03 solids pre-
cipitated. DSC analyses and microscopic examination provided
semiquantitative information on the identification of the solid
products.
Table 2-1 is a summary of the experiments conducted.
TABLE 2-1. SUMMARY OF EXPERIMENTAL CONDITIONS IN TESTS OF
MAGNESIUM SULFITE HYDRATE PRECIPITATION
FROM SCRUBBER-LIKE MEDIA1-
Test No.
Initial Sulfite
Concentration (gmole/S,)
Composition of 1 gram
of Seed Crystals
3-2
0.316
MgS03•3H2O
3-3
0.475
MgS03 *6H20
3-4
0.316
MgS03•3H20
3-5
0.367
MgS03 *6H20
3-6
0.374
MgS03*3H20 (0.5g) and
MgS03«6H20 (0.5g)
3-7
0.358
MgS03*3H20 (0.5g) and
MgSO3 *6H20 (0.5)
Tests were done at 55°C and pH 6 in solutions containing
approximately 20 wt 7. MgSOu. Initial magnesium concentra-
tion varied from 2.01 to 2.04 gmole/A.
242

-------
SECTION 3
DATA ANALYSIS METHODS
For batch-solid/batch-liquid kinetics experiments, the
precipitation rate can be calculated from the amount of solid
material produced as a function of the time. The total number
of moles of solids produced is plotted versus reaction time.
The precipitation rate can be obtained from the slope of the
resulting curve. The precipitation rate, R, at time t is ex-
pressed in equation 3-1.
N = moles of solids produced
t « duration of the run in minutes, and
R « precipitation rate in mMoles/min at time t
The amount of precipitating material produced at time t
can be determined from the change of concentration of ions in
solution. That is, the change in the solution magnesium or sul-
fite ion concentration with time as determined by analytical
means must be equivalent to the number of moles of magnesium or
sulfite reacted to form the MgS03 hydrate product. The amount
of precipitated material produced at time t was calculated from
the following equation:
(3-1)
where,
N(t) ~ N(t ) " <(I0Nlt "	Vo1
Reactor
(3-2)
where,
243

-------
[Ion] = initial magnesium or sulfite ion
concentration in solution in
moles/£, and
Vol,
Reactor
= total reaction solution volume in
liters.
Since the magnesium concentration was much greater than
the sulfite concentration, the change in sulfite concentration
was used to calculate the amount of precipitating material.
It is useful to describe precipitation rates with an equa-
tion which expresses the rate as a function of the important
variables which determine it. This correlation of experimen-
tally determined rate data is useful because it provides a
means of calculating the rate in other systems. Using appro-
priate values for the variables, a precipitation rate can be
predicted for design purposes. Equipment sizes and residence
times can thus be established. Equation 3-3 shows an appropriate
form for a rate correlation.
This correlation form is based upon the idea that the precipita-
tion rate is proportional to the crystal surface area and a
driving force. If no nuclei are being formed and the McCabe AL
law holds, the first two terms may be combined.
fesH liirl bar}
(3-3a)
(3-3b)
244

-------
k = k*B
(3-4)
k*A = k*W B(W/W J2/3	(3-5a)
s	o o
k*A_ - kW (W/W )2/3	(3-5b)
s	o o
In equation 3-5, W is moles of solids formed, and B is the sur-
face area factor in units of cm2 crystal surface area per mole
of initial seed.
The driving force is a function of the activities of ions
in solution. It takes into account the difference between the
activities at equilibrium and those at actual conditions. It
may be expressed in terras of the relative saturation, r.
r6 - a ¦.a a6 /Ksp6	(3-6a)
Mg SO3 H20
- ap6/Ksp6	(3-6b)
r3 « ap3/Ksp3	(3-7)
The moles of solids formed are normalized by dividing by
the initial number of moles of seed, W .
o
p - W/WQ	(3-8)
Equation 3-3 may be rewritten in terms of the new vari-
ables, and f	(3_9)
245

-------
Simplifying the derivative terms gives equation 3-10.
" 3«d(r)	(3"10)
Equation 3-10 shows that rates should be calculated based on a
173
graph of ^ ' versus reaction time.
The time scale for the reaction is such that it is not con-
venient to use a linear abscissa. A convenient logarithmic
scale defined such that the time origin is nonzero is given in
equation 3-11.
9 = ln(t+10)	(3-11)
The time derivative indicated in equation 3-11 can be ob-
tained in terms of 6 by the following relationship:
df - H ar	(3"12a)
" Tt+W 3F	(3-12b)
In an effort to determine a driving force term, activity
calculations were made using the Radian equilibrium model.
Based upon the solution composition data, the relative satura-
tions of the tri- and hexahydrates of MgS03 were calculated.
These calculations indicated that the system was approaching a
steady state relative saturation of about 2.5 rather than 1.0
as required by theory. Equilibrium solubility experiments at
high ionic strength confirm this order of magnitude discrepancy
between experimental and calculated values.
246

-------
These results indicate that the equilibrium calculations
made in these high ionic strength solutions are not accurate
enough to correlate the data generated in this work. An update
of either the theoretical basis of the calculations or the
parameters used is indicated.
Since the equilibrium model could not be used as an analy-
tical tool, it was not possible to calculate accurate driving
force expressions for use in the rate data correlation. As a
result, no attempt was made to calculate the rate constant for
the correlation form shown in equation 3-3. The correlation
form does show which variables are important in determining
precipitation rates. And it indicates the proper form in which
to plot rate data.
247

-------
SECTION 4
RESULTS AND DISCUSSION
The results of the precipitation rate experiments are
presented in Tables 4-1 through 4-6. The results of chemical
analyses of slurry liquor samples were used to calculate moles
solids precipitated. The calculation was based on the decrease
in sulfite concentration in the solution, since the change in
magnesium concentration was small compared to the total concen-
| |
tration. The total measured decrease in moles of Mg ion was
approximately equal to that of the S07 ion within experimental
1/3
error. The tables also give values for ip ' as defined in
Section 3. Examples of the results of DSC analyses of solid
samples are given in Figures 4-1 through 4-3. Figure 4-1 shows
the DSC scan for a pure sample of magnesium sulfite hexahydrate
and Figure 4-2 shows the DSC scan for a pure sample of the
trihydrate. As indicated in the figures, the hexahydrate peak
has a maximum at about 110°C and the trihydrate maximum occurs
at about 170°C. Figure 4-3 shows an example from the DSC
analysis of solids precipitated during Experiment No. 3-5.
The solid sample was collected after 470 minutes. Figure 4-3
indicates that both the hexahydrate and the trihydrate are
present in the solid sample.
Examples of electron micrographs of pure hydrates and the
same sample from experiment 3-5 are shown in Figures 4-4 through
4-6. Figure 4-4 shows rhombic hexahydrate crystals at 300X
magnification. The much smaller trihydrate crystals are shown
in Figure 4-5 at 1000X magnification. The electron micrograph,
in Figure 4-6, obtained at 1000X shows small trihydrate cry-
stals growing on the surface of the much larger hexahydrate
crystals.
248

-------
TABLE 4.1. RESULTS OF MAGNESIUM SULFITE HYDRATE
PRECIPITATION EXPERIMENT NO. 3-2*

Results of Chemical Analyses
of Slurry Liquor at Indicated
Sample Time
•kk
Moles

Sample
Time
w ++
Mg
SO
3
pH
•kick
*1/3
emole
emole
emole
emole

Solids
(Min)
£
kg water
a
kg water

Precipitated
0
2.02
2.07
0.316
0.324
6.4
0
1.00
60
1.94
1.99
0.303
0.310
6.4
0.037
1.92
180
1.94
1.99
0.304
0.311
6.4
0.034
1.88
260
1.90
1.94
0.295
0.301
6.4
0.058
2.20
1260
1.88
1.92
0.278
0.283
6.4
0.104
2.62
1700
1.91
1.95
0.275
0.281
6.4
0.110
2.67
3150
1.96
2.00
0.271
0.277
6.4
0.119
2.74
7000
1.95
1.99
0.248
0.253
6.4
0.176
3.10
8860
1.93
1.96
0.230
0.234
6.4
0.220
3.33
11500
1.93
1.92
0.228
0.232
6.4
0.221
3.34
* Experiment was conducted at 55°C using 1 gram of magnesium
sulfite trihydrate seed crystals. The initial sodium ion
concentration was 0.648 gmole/kg water and the initial sul-
fate ion concentration was 2.1 gmoles/kg water. The density
of the liquid was 1.260g/ml.
** The moles solids precipitated were calculated from the
number of moles of sulfite leaving solution.
*** xh = total moles of solids
initial moles of seed crystals
249

-------
TABLE 4-2. RESULTS OF MAGNESIUM SULFITE HYDRATE
PRECIPITATION EXPERIMENT NO. 3-3*

Results of Chemical Analyses
of Slurry Liquor at Indicated
Sample Time


Sample
Time
(Min)
mr
Mg
sol
pH
icif
Moles
Solids
Precipitated

gmole
I
gmole
kg water
Rmole
%
£mole
kg water

,1/3
0
2.04
2.14
0.475
0.498
6.41
0
1.00
10
1.98
2.06
0.432
0.450
6.48
0.128
3.04
20
1.93
2.00
0.401
0.416
6.40
0.218
3.62
30
1.95
2.02
0.398
0.413
6.36
0.224
3.64
40
1.93
2.00
0.384
0.398
6.30
0.260
3.83
60
1.88
1.94
0.366
0.378
6.25
0.306
4.04
150
1.87
1.93
0.355
0.367
6.27
0.332
4.14
200
1.88
1.94
0.358
0.370
6.30
0.318
4.09
1140
1.90
1.96
0.348
0.359
6.30
0.339
4.18
* Experiment was conducted at 55°C using 1 gram of magnesium
sulfite hexahydrate seed crystals. The initial sodium ion
concentration was 0.996 gmole/kg H20 and the initial sulfate
ion concentration was 2.17 gmole/kg H20. The density of the
liquid was 1.260 g/ml.
** The moles solids precipitated were calculated from the
number of moles of sulfite leaving solution.
** ib = total moles of solids
v initial moles of seed crystals
250

-------
TABLE 4-3. RESULTS OF MAGNESIUM SULFITE HYDRATE
PRECIPITATION EXPERIMENT NO. 3-4*

Results of Chemical Analyses
of Slurry Liquor at Indicated
Sample Time


Sample
Time
(Min)
MR"14"
SO 3
PH
•kit
Moles
Solids
Precipitated

gmole
£
gmole
kg water
gmole
Si
gmole
kg water


0
2.02
2.07
0.316
0.324
6.4
0
1.00
10
2.02
2.07
0.313
0.321
6.4
0.009
1.43
20
2.02
2.07
0.313
0.321
6.4
0.009
1.43
40
1.98
2.03
0.305
0.312
6.4
0.032
1.98
60
1.87
1.91
0.291
0.297
6.4
0.072
2.53
150
1.84
1.87
0.280
0.285
6.4
0.101
2.82
280
0
CO
1—1
1.83
0.275
0.280
6.4
0.113
2.92
1260
1.81
1.83
0.216
0.219
6.4
0.271
3.88
1740
1.82
1.84
0.196
0.198
6.4
0.319
4.10
3150
1.86
1.88
0.166
0.168
6.4
0.392
4.38
7000
1.87
1.89
0.161
0.163
6.4
0.397
4.40
8860
1.87
1.89
0.165
0.167
6.4
0.379
4.33
* Experiment was conducted at 55°C using 1 gram of magnesium
sulfite trihydrate seed crystals. The initial sodium ion
concentration was 0.648 gmole/kg H2O and the initial sulfate
ion concentration was 2.10 gmole/kg H20. The density of the
liquid was 1.260 g/ml.
** The moles solids precipitated were calculated from the number
of moles of sulfite leaving solution.
•kit
iji - total moles of solids
initial moles of seed crystals
251

-------
TABLE 4-4. RESULTS OF MAGNESIUM SULFITE HYDRATE
PRECIPITATION EXPERIMENT NO. 3-5*

Results of Chemical Analyses
of Slurry Liquor at Indicated
Sample Time


Sample
Time
(Min)
W ++
Mg
SO
1
PH
Moles
Solids
Precipitated

gmole
l
gmole
kg water
gmole
I
gmole
kg water


0
2.01
2.07
0.367
0.378
6.2
0
1.00
10
2.01
2.07
0.349
0.359
6.2
0.054
2. 31
20
1.99
2.05
0.350
0.360
6.2
0.050
2.26
30
1.99
2.05
0.349
0.359
6.2
0.052
2.29
40
1.97
2.03
0.347
0.357
6.2
0.057
2.36
60
1.99
2.05
0.351
0.361
6.2
0.045
2.19
80
1.99
2.05
0. 346
0.356
6.2
0.058
2.36
100
1.90
1.95
0.327
0.335
6.2
0.108
2.88
120
1.97
2.02
0.342
0. 352
6.2
0.066
2.47
150
1.92
1.97
0.332
0.341
6.2
0.092
2.73
180
1.91
1.96
0.340
0.349
6.2
0.069
2.50
270
2.00
2.06
0.344
0.354
6.2
0.058
2.36
330
1.93
1.98
0.324
0. 332
6.2
0.106
2.86
390
1.93
1.98
0.324
0.332
6.2
0.103
2.84
470
1.98
2.03
0.330
0. 339
6.2
0.087
2.69
1400
1.86
1.89
0.246
0.250
6.2
0.277
3.91
3200
1.83
1.85
0.180
0.182
6.2
0.417
4.47
* Experiment was conducted at 55°C using 1 grain of magnesium
sulfite hexahydrate seed crystals. The initial sodium ion
concentration was 0.756 gmole/kg H20 and the initial sulfate
ion concentration was 2.10 gmole/kg H20. The density of the
liquid was 1.260 g/tnl.
** The moles solids precipitated were calculated from the number
of moles of sulfite leaving solution.
•fc-fc
tp =, total moles of solids
initial moles of seed crystals
252

-------
TABLE 4-5. "RESULTS OF MAGNESIUM SULFITE HYDRATE
PRECIPITATION EXPERIMENT NO. 3-6*
Sample
Time
(Min)
Results of Chemical Analyses
of Slurry Liquor at Indicated
Sample Time
VrVf
Moles
Solids
Precipitated


soT
PH '
KQiole stmole
t kg water
gmole gmole
S, kg water

0
2.01 2.08
0.374 0.386
5.9
0
1.00
10
1.94 2.00
0.364 0.375
5.9
0.030
1.86
20
1.98 2.04
0.375 0.387
5.9
0.001
-
30
1.96 2.02
0.363 0.374
5.9
0.033
1.91
40
1.94 2.00
0.358 0.369
5.9
0.048
2.13
50
1.95 2.01
0.360 0.371
5.9
0.042
2.04
60
2.00 2.06
0.370 0.382
5.9
0.012
1.46
80
1.99 2.05
0.367 0.379
5.9
0.021
1.69
100
1.99 2.05
0.365 0.377
5.9
0.027
1.81
120
1.99 2.05
0.363 0.374
5.9
0.033
1.91
180
1.98 2.04
0.358 0.369
5.9
0.048
2.13
240
1.95 2.01
0.354 0.365
5.9
0.060
2.25
300
1.95 2.01
0.357 0.368
5.9
0.051
2.17
390
1.95 2.01
0.345 0.355
5.9
0.087
2.56
4260
1.90 1.93
0.204 0.207
5.9
0.510
4.54
* Experiment was conducted at 55°C using 0.5g each of magnesium
sulfite trihydrate and hexahydrate seed crystals. The initial
sodium ion concentration was 0.772 gmole/kg H2O and the initial
sulfate ion concentration was 2.11 gmole/kg H20. The density
of the liquid was 1.260 g/ml.
** The moles solids precipitated were calculated from the number
of moles of sulfite leaving solution.
^ , total moles of solids
initial moles of seed crystals
253

-------

TABLE 4-6
. RESULTS OF MAGNESIUM SULFITE
HYDRATE PRECIPITATION EXPERIMENT NO. 3-7*


Results of Chemical Analyses
of Slurry Liquor at Indicated
Sample Time



Sample
Time
Mr'"'
S03~
PH
**
Moles
Solids

gmole
gmole
gmole
gmole

***
(Min)
a
kg water
I
kg water

Precipitated
^1/3
0
2.03
2.10
0.358
0.370
6.0
0.000
1.00
10
1.88
1.93
0.340
0.349
6.0
0.054
2.21
20
2.00
2.06
0.337
0.347
6.0
0.063
2.32
30
1.94
1.99
0.332
0.341
6.0
0.078
2.47
40
1.91
1.96
0.322
0.330
6.0
0.108
2.74
60
1.89
1.94
0.313
0.321
6.0
0.135
2.94
80
1.93
1.98
0.316
0.324
6.0
0.126
2.88
100
1.93
1.98
0.321
0.330
6.0
0.111
2.76
120
1.87
1.92
0.316
0.324
6.0
0.126
2.88
150
1.93
1.98
0.318
0.326
6.0
0.120
2.83
180
1.96
2.01
0.326
0.335
6.0
0.096
2.64
240
1.99
2.05
0.320
0.329
6.0
0.114
2.79
320
2.00
2.06
0.330
0.340
6.0
0.084
2.53
380
1.91
1.96
0.317
0.325
6.0
0.123
2.86
1400
1.95
1.99
0.272
0.278
6.0
0.258
3.63
* Experiment was conducted at 55°C using 0.5g each of magnesium sulfite trihydrate and hexahydrate
seed crystals. The initial sodium ion concentration was 0.740 gmole/kg H2O and the initial
sulfate ion concentration was 2.13 gmole/kg H2O. The density of the liquid was 1.260 g/ml.
** The moles solids precipitated were calculated from the number of moles of sulfite leaving solution.
*** ^ _ total moles of solids
initial moles of seed crystals

-------
X/"
Run No: 2
Date: 3-18-76
Operator: RES
Sample: MgSC>3 • 3H20
Covered Pan
DTA-DSC
Scale, °C/in: 10
Weight, mg: 0.46
Reference: Empty Pan
20 AO 60 80 100 120 140 160 180 200
Tempe ra ture, °C
Figure 4-1. Results of DSC Analysis of a Pure
Sample of MgS03*3H20

-------
I
Run No: 3
Date: 3-18-76
Operator: RES
Sample: MgS03*3H20
Covered Pan
Flow Rate: Start 27°C
DTA-DSC
Scale, °C/in: 10
Weight, mg: 0.83
Reference: Empty Pan
» i		 i « i			. . i .	i	•	i	 i - ¦ . ¦
20 40 60 80 100 120 140 160 180 200
Temperature, °C
Figure 4-2. Results of DSC Analysis of a Pure
Sample of MgS03*3H20

-------
80
70 ¦
60
SO
>
t-
U)
z
tu
40
Ui
>
30
eo
80
100
120
160
140
180
TEMP. t*CJ
Figure 4-3. Results of DSC Analysis of Solids Precipitated
After 470 Minutes in Experiment No. 3-5.
257

-------
FIGURE 4-4.
ELECTRON MICROGRAPH 300X
OF MgS03'6H20 CRYSTALS (RHOMBIC)
NUCLEATED AT 28° C
258

-------
FIGURE 4-5.
ELECTRON MICROGRAPH 1000X OF
MgS03. 3H20 CRYSTALS NUCLEATED
AND GROWN OVERNIGHT AT 65°C
259

-------
FIGURE 4-6.
ELECTRON MICROGRAPH (1000X)
FROM R 3-5 AT T= 470 MINS
TRI CRYSTALS GROWING ON
LARGER HEXA CRYSTALS.
260

-------
DSC scans for solid samples obtained during the experiments
described in Tables 4-1 through 4-6 are included in the appendix.
These analyses, and the results of microscopic analyses, were
used for qualitative identification of the solids produced
during each precipitation experiment.
The precipitation rate data from each experiment are
l/o
plotted in the form ip ' versus 6 in Figures 4-7 through 4-12.
These figures also indicate the composition of the solid pro-
ducts. The approximate fraction of trihydrate present at each
sample point is shown at the top of the graphs. Observations
and comments about what occurred during each experiment are
summarized in the following paragraphs.
Experiment 3-2, Trihydrate Seed Crystals
(Table 4-1, Figure 4-7, Figures A-l and A-2
The solid products from this experiment consisted solely
of the trihydrate. At the experimental conditions of 0.316
gmole/£ initial sulfite concentration and trihydrate seed
crystals it is apparent that nucleation of the hexahydrate did
not occur. Primary nucleation of hexahydrate was observed at
55°C in 20 wt % MgSO^ solutions at a sulfite concentration of
about 0.55 gmole/kg H20. In secondary nucleation experiments
with trihydrate seed crystals at 55°C hexahydrate nucleation
was not observed.
The trihydrate exhibited a constant precipitation rate.
Since a dramatic increase in the initial crystal growth rate
was not observed it appears as if secondary trihydrate nuclea-
tion did not occur. This observation is consistent with the
results of secondary nucleation experiments reported in
Technical Note 200-045-54-05.
261

-------
Fraction
Trihydrate
ro
*0
4.0
^20 wt % MgSOij solution
T = 55 °C
Initial conditions:
2.0
SO3 = 0.316 fmole/K,
Ip"1"* = 2.02 gmole/S,
1.0
10
100
1000
10,000
¦ 0 = 10 + Sample Time (minutes)
Figure 4-7. Results of MgS03 Hydrate Precipitation
Rate Experiment 3-2 Employing Trihydrate
Seed Crystals

-------
Experiment 3-3, Hexahydrate Seed Crystals
(Table 4-2, Figure 4-8, Figures A-3 through A-7)
A high initial sulfite concentration of 0.475 gmole/Jl was
employed in this experiment. Hexahydrate nucleation was ob-
served visually in the first 20 minutes. These results occurred
at approximately the same solution conditions as those reported
in previous secondary nucleation experiments at 59.2°C. (Techni-
cal Note 2-0-045-54-05). In addition, the graph of	vs. 9
in Figure 4-8 shows an initial high rate of solids growth which
is consistent with the observation of nucleation. The results
of DSC analysis indicated the product was pure hexahydrate at
the end of 200 minutes. Formation of trihydrate crystals was
not observed.
Experiment 3-4, Trihydrate Seed Crystals
(Table 4-3, Figure 4-9, Figures A-8 and A-9)
This test was done under the same conditions as experiment
3-2. Again the products consisted solely of the trihydrate and
a relatively constant growth rate occurred. The precipitation
rate was faster than that in Run 3-2, although the same initial
sulfite concentration was employed and the driving forces should
be the same in the two experiments. The faster growth rate
could be the result of seed crystal surface area effects.
Smaller seed crystals provide a larger surface area per volume.
The seed crystal preparation method employed could have re-
sulted in some variations in crystal size.
It is apparent from Figure 4-9 that the system approached
equilibrium at about 10,000 minutes.
263

-------
Fraction
Trihydrate
4.0
3.0
2.0
¦i	*—
_L

100
1000
0 = 10 + Sample Time (minutes)
^20 wt 7, MgSOi, solution
T = 55°C
Initial Conditions:
pH = 6.41
S03 = 0.475 gmole/Ji
Mg"1"1" = 2.04 gmole/i.
Figure 4-8. Results of MgS03 Hydrate Precipitation Rate
Experiment 3-3 Employing Hexahydrate Seed Crystals

-------
Fraction
Trihydrate
^20 wt % MgSOi* solution
T = 55°C
Initial Conditions:
pH « 6.4
SO3 = 0.316 gmole/Jl
Mg"*"*" « 2.02 gmole/Jl
10
100
1000
10,000
"0 « 10 + Sample Time (minutes)
Figure 4-9. Results of MgS03 Hydrate Precipitation Rate
Experiment 3-4 Employing Trihydrate Seed Crystals

-------
Experiment 3-5, Hexahydrate Seed Crystals
(Table 4-4, Figure 4-10, Figures A-10 through A-21)
This experiment employed a lower initial sulfite concen-
tration than experiment 3-3, which also employed hexahydrate
seed crystals. Hexahydrate nucleation was not apparent from
visual observations, and it is not clearly indicated in Figure
4-10.
The results of DSC analyses of solid samples indicated
initial formation of the hexahydrate followed by complete con-
version to the trihydrate. In the first part of the run, an
increase in total solids occurred which is attributed to the
precipitation of hexahydrate. This is confirmed by DSC analyses
which show almost pure hexahydrate up to about 300 minutes. It
is probable that at the end of 100 minutes the solution is
saturated with respect to hexahydrate precipitation. In the
intermediate period up to 500 minutes the amount of solids
present changes relatively little while the fraction of tri-
hydrate is increasing. These results indicate that the hexa-
hydrate is dissolving and the trihydrate is precipitating.
After about 400 minutes an increase in the fraction of trihy-
drate occurs. The trihydrate precipitation rate is initially
slow because there is very little trihydrate crystal surface
area available. As the trihydrate surface area increases,
an increase in the rate of trihydrate precipitation is ob-
served. This increase is apparent after 400 minutes in Figure
4-10.
Trihydrate crystals could have occurred as a result of
nucleation, since the solution was supersaturated with respect
to trihydrate for some period of time. However, clear evidence
of secondary trihydrate nucleation has not been obtained in
previous experiments. It is also possible that the occurrence
266

-------
Fraction
Trihydrate
m
PO
4.0
u

•H
a.20 wt % MgSOi* solution
T = 55°C
Initial Conditions:
3.0
tn
S3
2.0
SO3 = 0.367 gmole/A
[g^ = 2.01 gmole/&
1.0
10	100	1000
9 = 10 + Sample Time (minutes)
Figure 4-10.
Results of MgS03 Hydrate Precipitation Rate
Experiment 3-5 Employing Hexahydrate
Seed Crystals

-------
of trihydrate crystals can be attributed to seed crystal char-
acteristics. Growth of trihydrate crystals on the surface of
hexahydrate crystals was observed as was shown in Figure 4-6.
Crystal surface imprefections or the presence of small amounts
of trihydrate seeds could account for initiation of trihydrate
precipitation. DSC scans of samples taken very early in the
test showed very small amounts of trihydrate crystals. It is
not clear from the experimental results what initiated the
formation of trihydrate crystals in tests employing hexahydrate
seed crystals.
Experiment 3-6, Mixed Tri- and Hexahydrate Seed Crystals
(Table 4-5, Figure 4-11, Figures A-22 through A-28)
The results of this mixed seed crystal test at an initial
sulfite concentration of 0.374 gmole/£ can be explained on the
same basis as those of experiment 3-5. Again, immediate pre-
cipitation of the hexahydrate followed by hexahydrate dissolu-
tion and trihydrate precipitation were observed. At the end of
4000 minutes all the hexahydrate had dissolved and the solid
product was pure trihydrate.
Experiment 3-7, Mixed Tri- and Hexahydrate Seed Crystals
(Table 4-6, Figure 4-12, Figures A-29 through A-36)
This experiment was done under the same conditions as
experiment 3-6, except the initial sulfite and magnesium con-
centrations were slightly different. The same qualitative re-
sults vcere observed, but the initial rate of solids production
was higher. The difference in precipitation rate can be attri-
buted to either seed crystal surface area effects or a differ-
ence in driving force terms.
268

-------
Fraction
Trihydrate
to
ON
NO
m
ID
•o
CO
OS
•o
•H
c
10
100
1000	10,000
10 + Sample Time (minutes)
o,20 wt % MgSO,, solution
T = 55°C
Initial conditions:
pH = 5.9
S03 = 0.374 gmole/£
Mg^ =2.01 gmoleM
Figure 4-11. Results of MgS03 Hydrate Precipitation
Rate Experiment 3-6 Employing a Mixture
of Tri- and Hexahydrate Seed Crystals

-------
Fraction
Trihydrate
5.0
•o
4.0
4J
^20 wt 7. MgSOit Solution
T = 55°C
Initial conditions:
3.0
ci
2.0
S03 = 0.358 gmole/£,
Mg++ =2.03 gmole/S.
1.0
10	100	1000	10,000
0 = 10 + Sample Time (minutes)
Figure 4-12. Results of MgS03 Hydrate Precipitation
Rate Experiment 3-7 Employing a Mixture
of Tri- and Hexahydrate Seed Crystals

-------
SECTION 5
CONCLUSIONS
The results of these precipitation rate studies provide
an accurate qualitative explanation of what occurs during MgS03
hydrate precipitation under conditions characteristic of the
magnesium oxide wet scrubbing process. Data are available for
obtaining a quantitative rate correlation; however, some re-
finement of computational tools will be required in order to
calculate accurate driving force terms.
The results for mixed hydrate precipitation in scrubber-
like media are consistent with those for precipitation of indi-
vidual hydrates in dilute solutions and with tests of primary
and secondary nucleation in scrubber-like media. Hexahydrate
nucleation and precipitation rates are faster than those of the
trihydrate. The trihydrate is the thermodynamically stable
form. Clear evidence of trihydrate nucleation is not apparent.
These results provide an adeuqate basis for the design of a
system to produce either hydrate as a solid product.
271

-------
APPENDIX A
TO TECHNICAL NOTE
200-045-54-03a
FIGURES A-l THROUGH A-36
RESULTS OF DSC ANALYSIS OF
SOLID SAMPLES FROM PRECIPITATION
RATE EXPERIMENTS IN SCRUBBER-LIKE MEDIA
272

-------
=3

3=

Figure A-l
Results of DSC Analysis of
Solid Sample Taken at
60 minutes In Experiment 3-2
"¦MB
US8
LlUiUtl

273

-------
'A I
O"
Figure A-2
Results of DSC Analysis of
Solid Sample taken at
3150 minutes In Experiment 3-2
Cjr
274

-------








—4




































1









	







	














1






—




^
——



	
-	


	



.
_r	;:::
	









	
— "r	
	


——i	




„




		
....






— -
,,



... .1


	


Figure A-3

r-n


Solid Samp la Talcen at
10 minutes in Experiment 3-3
T=r=
9
muuiuiu
U1UUI
utuiu
275

-------
Figure A-4
Results of DSC Analysis of
Solid Sample Taken at
20 minutes in Experiment 3-3
¦XT
Ttapat«tur11, *C
mtuuiiiumitnammimu
jmua
276

-------
Figure A-5
Results of DSC Analysis of
Solid Sample taken at
60 minutes in Experiment 3-3
&SL
o
Ti mp«r iture, *C
IIHI.lll IIHIIlll IIIIIIlL llll.lll lllllll*
ll ll.ll.lU IIIIIIIM lllllttl
am
277

-------
Figure A-6
Results of DSC Analysis of
Solid Sample taken at
150 minutes in Experiment 3-3
i |	T*m wnt >r«
W		'*"* »"'^" 	^
unit it mm
imnfai «md« imnki .hh^i
278

-------
! I
-r=


tmt
——	Figure A-7
— Results of DSC Analysis of
Solid. Sample taken ac
200 minutes in Experiment 3-3


00CZ-9S0 ON ibVHD
279

-------
Figure A-8
Results of DSC toialysis of
Solid Sample taken at
60 minucea In Experiment 3-4
142
280

-------
Figure A-9
Results of DSC Analysis of
Solid Sample taken at
1260 minutes In Experiment 3-4
T« aperi itura , *C
IPWMHW
tm
281

-------
Figure A-10
Results of DSC Analysis of
Solid Sample taken at
20 minutes in Experiment 3-5
282

-------
FlgMT* A-ll
lUsulca of DSC An«lyal« of
Solid S*npl« eakan *t
60 minutu in Exp«riatne 3-5

191 pttl
uuimmumi
mmnittinm
ntouuuiuii
wmii wimii
283

-------
Figure A-12
Results of DSC Analysis of
Solid Sample taken at
100 minutes In Experiment 3-5
iart i.tHw nfci.iH
I	IUWUA
iimitti m
284

-------

i- ¦ -1 ¦
•!	i~
! .1
-XI-
Figure A-13
Results of DSC Analysis of
Solid Sample taken at
120 minutes in Experiment 3-5
-•+
285

-------
Figure A-14
Results of DSC Analysis of
Solid Sample taken at
150 minutes in Exparlmenc 3-5


jBaaapu^aaritTtitmringifc
286

-------
Figure A-15
Results of DSC Analysis of
Solid Sample taken at
180 minutes In Experiment 3-5
— T« ip*r» s««», *C-	
287

-------
ijut' "nig*.	j.mulu,iy"
288

-------
: ... i
r	¦
Figure A-17
Results of DSC Analysis of
Solid Sample Taken at
330 minutes in Experiment 3-5
289

-------
Figure A-18
Results of DSC Analysis of
Solid Sample Taken at
390 minutes in Experiment 3-5
o
160-
80

290

-------
Figure A-19
Results of DSC Analysis of
Solid Sample Taken at
470 minutes In Experiment 3-5
10
120
iMfiHtun, -TC-
r w
JUUI
291

-------
Figure A-20
Results of DSC Analysis of
Solid Sample Taken at
1400 minutes in Experiment 3-5
155

lini iiny^ii
llUUUl
• uw
292

-------
Figure A-21
Results of DSC Analysis of
Solid Sanple taken at
3200 minutes in Experiment 3-5
16?
- 				_ Ta^wMur^ -Xj - -
Jim. h.L.h'.mLh hi	|mh ygipi»Lj.^ i^, J..^|.uL.iT^r
293

-------
Figure A-22
Resales of DSC Analysis of
Solid Sample taken ac
20 minutes In Experiment 3-6
U 111
""1"TTHII|l|||l""'r'TTT
T«ap*r«tura, *C

294

-------
. |.....
L.._l I. !


I-.
• j i
r •




i

.
i . ¦ . . - r

! r i f 1 -
-O ,
: 4. . -- 1
- ¦ ; i
t . v
I— • • 			
}	 i - • ~ - -	Figure A-23

I

.
Raaults of DSC Analysis of
Solid Sample taken ac
50 mlnuCas In Experiment 3-6
! i
\ 1
....I

1


. | • • 1 - .
1 1 1
-t	i :



r -i
i ¦ • ¦
i
1 i
CO
:
.... 4 ... 1 -¦ - .
. 1
; - ; i
I ¦ : i
iii''
1 i ; ! • ; ¦
1 i
' 1 "
• ; i ..i !.. .
*r T r : - • .
i • i; : ;





1 ! ¦ •
: t

i i
¦¦I "h
- -\ -; ¦; ;	-t :
. < :

.....


-

		
V\


— 4-_. L.. -1— - }-. ¦ - i-... --.
. .1 ...j :

--







	4- —
-pr.-'-h i

..:t




	

~."


.....} j. r-

	—
a> T
-
j" "1 ' !








- •
•
.
1 1 '
-¦ i—-1— :












r -.. .
:




"""

..».





-	i ¦ -¦
• i







;



.......


:Lz



... .j .
\
1
t": i.: rT"
! !
1
i
i

.. ...
—
.....
- - 7

s\




| j







\


-1 : ...... j.. ;
11^ 1 !


.

.:~r

s.

Nj VO-
-\r l - i
-
-
	L	
—f





I
- i—-}• - I




120	160
Ta^araeura, *C
295

-------
.
; 		 ; *	¦
» 1 f

• 1 -

t - i t r • •

.. i
t
i-

• ' i
! .. . i-

j j ..L
Results of DSC Analysis
Solid Sample taken at
60 mlnut«8 in Experlmen
°£ i ' i
V - * i .
t 3-6 : i 1

¦ i
: i

U !


-
--
• r
: ! \
, ; "r":.r
; 1

i i


• : - t

r!
.u.
j- ¦ I
i.;- : ¦ - I
— - -j" ' - t ' - ' *

! ' '
! P I"
i
-
rI	ri- ; - i


... .... ..t. . -|	
1 ! . 1 .
: !-"	i"'"
.. i..,.
-••• •
.... .j.
"- j ¦" : ¦¦¦ ; "


' f -
TP
pi
t t -..P j. _ !
! ! : i i

- f
t-
" ! "
- i -•
¦ i

1
. . ... ... _
- - - r v- ¦
: t ! . ..
j ¦ i- - i i
•r ! ¦ : ;


t- !
i ,
i (


1 1 -1
!¦ -j i ¦ j

; t
i !
.. 1_ t
/ ! \
|. ..
1 ; i

	; ,i. )/ . , v
: i 1 \
t - f--
; i
r
i i . i
" ¦"¦P-L: I "" " -
—! - . ; . /
¦¦ p ' I ; 11	t
. /
L I 1 |
j ; r - 1 • - i - f-
i 1 /
: ':, . j \ /
¦ \ 1 1/
"...
I i :t 1:1-!' i ¦ f -
i J ! j


\ i • j
j ;
I"
i
i


t f
T r"
-p ^
S ' itvO_ ¦'

4	
4-i

M9-
- 1 i -

)iiiiitfLiw^iiiji
-------
Figure A-25
Results of DSC Analysis of
Solid Sample Taken at
180 minutes In Experiment 3-6
120
Teayereture, *C
160
297

-------
Figure A-2 6
Result* of DSC Analysis of
Solid Simple Taken at		j.
300 minutes in Experiment 3-6
Teaparature, *C
298

-------

















































































































- —1










Figure A-27
Results of DSC Analysis of
Solid Sample Taken at
390 minutes in Experiment 3-6














































	











——














































































































































































































































































































—y
"\













J—
-V













i—
=V












=±
—













-h


t—











/—
—
















=fc





























—^
	







































«n mil.II yum. inii.ii Jin II
80	120	160
Temperature, °C	02-2308-•
-299-

-------



















































































































































































	





































































Fieure A-28








of







Results or DSC Analysis








Solid Sample Taken at















L-)hC\

-h







































































































































































































































. _J ..











































































1




















































































































































































































































[













1













f













\













I













1













[















































































































































































¦v














A














\




























\














>















I





























\














V














\






























V





























"" V
















"¦


























hum:
IIII1I1U
llllillil
mimir
ItllUHL
11! run
IlllllH
llllillil
llllllll
llllllll
llllllll
llllllll
llllltll
Ill llll
llllllll







80	120	160
Temperature, °C
-300-

-------
> . —
.. t - -
j
- • ¦
i
; j ' j
: i '•
p:,
L	t
Figure A-29
Results of DSC Analysis of
Solid Sample Taken at
10 minutes In Experiment 3-7
- r
"~j
i
TV-TIl
r
	
t-
-t	
L. ....
. . 1
	1
J	
_
:
m
Kl(T
Trapcratur*, *C
301

-------
Figure A-30
Results of DSC Analysis of
Solid Sample Taken at
30 minutes In Experiment 3-7

120
160
•
Tmmytmtwn, C
302

-------

-pr:
f —
1—
£T	
---t-


-g—
Figure A-31
Results of DSC Analysis of
Solid Sample Taken at
40 minutes In Experiment 3-7
n» latfflmiwnJmua
Tmp«racur«, *C
303

-------
zzt

—r...
- -I
Figure A-32
Results of DSC Analysis of
Solid Sample Taken ac
SO minuces in Experiment 3-7
T«ap*Tarur*f *c
r »
304

-------
Figure A-33
RmuICs of DSC Analysis of Solid
Sample Taken at
150 mlnutea in Experiment 3-7
tiumsmuim
120
T«9*raeur«, *C
sum
305

-------
: "T" ! 	r ~-f- t ~T"'~r i "
¦- i-	i- • 1 i !
¦ ¦ ..,i 	 , r — — —
! t t ' 1
t r ¦ ; r ¦ :
...
1 ' j- -L ' —|
	.	i	4-	1	-j			j

l_ ... ... , ,
	— -| ¦ u 	f— --j-
... !
. . ; - » - - ; • | • :
i ! 1 1 I
. . . .. - -i
T - - - •' -• T ' . t
|
: t-
.... .. . l ¦
— _L Figure A-34 I-.-- 1 [	J
' Raaules of DSC Analysis of Solid ] ( ' "" v
240 minutaa in Experiment 3-7
80	120	160 r- >
Tasparaeura, *C
306

-------
80

-------

r—-—r	:	'
-------h—;¦

„ r_"—
. .

: - : i i
¦t 1 ; t i » • f * •
	 		 		i " 7	 ' * t • "
1 ! 	 * ' ;
; . : i" :
t .! t .1. . .L..J
- : - t .1 :
Figure A-36	¦ i- 	I-
Results of DSC Analysis of Solid	"T .)
Sample Taken at	| I !
1400 minutes in Experiment 3-7	1
126
Teopereture, "C
1M
308

-------
Technical Note 200-045-54-04
HOMOGENEOUS NUCLEATION OF
MgSO 3 HYDRATES IN SCRUBBER-LIKE
MEDIA
Prepared by:
I
J, L. Skloss
309

-------
CONTENTS
1.	Introduction	312
2.	Experimental Method for Homogeneous Nucleation. . . 313
3.	Results	315
4.	Discussion	320
5.	Conclusion	335
FIGURES
Number	Page
3-1 DSC Scans of Magnesium Sulfite Solids Nucleated in
Scrubber-Like Media at Various Temperatures .... • 317
3-2	Homogensous Nucleation from 20% MgSOi, Solutions and
and pH 6	319
4-1	X-Ray Diffraction Pattern of MgSO 3*6^0 Crystals
Nucleated at 28°C	321
4-2 X-Ray Diffraction Pattern of MgS03 *61120 Crystals
Nucleated at 50°C	321
4-3 X-Ray Diffraction Pattern of MgS03*3H20 Crystals
Nucleated at 69°C	323
4-4 Photomicrograph 125x of MgS03»6H20 Crystals (Rhombic)
Nucleated at 28°C	325
4-5 Photomicrograph 125x of MgS03*6H20 Crystals (Hexagonal)
Nucleated at 50°C	326
4-6 Photomicrograph 250x of MgS03*3H20 Crystals Nucleated
and Grown Overnight at 65 °C	327
4-7 Electron Micrograph 300x of MgS03*6H20 Crystals
(Rhombic) Nucleated at 288C 	 328
4-8 Electron Micrograph lOOx of MgS03»6H20 Crystals
(Hexagonal) Nucleated at 50®C		 . 329
4-9 Electron Micrograph lOOOx of MgS03*3H20 Crystals
Nucleated at 69°C	, . . , .330
310

-------
FIGURES (Continued)
Number	Page
4-10 Electron Micrograph lOOOx of MgS03*3H20 Crystals
Nucleated and Grown Overnight at 65 °C	331
4-11 Infrared Spectra of Magnesium Sulfite Nucleated
at 50°C and 69°C	334
TABLES
Number	Page
3-1	Solution Compositions Observed to Produce Homog-
eneous Nucleation at Various Temperatures	316
4-1	X-Ray Diffraction Data for MgS03*6H20	322
4-2 X-Ray Diffraction Data for MgS03*3H20. 		324
311

-------
SECTION 1
INTRODUCTION
The magnesium oxide process removes sulfur dioxide from
the flue gas by the precipitation of magnesium sulfite solids.
This precipitation occurs whenever the solubility product of
magnesium sulfite is exceeded if seed crystals are present. In
the absence of nuclei, metastable solutions of magnesium sul-
fite can be formed. Supersaturation can be increased only to
the limit whereby the magnesium sulfite solids are precipitated
spontaneously. In clear solutions this induced precipitation
is called primary nucleation. The purpose of this study was
to determine the compositions required to spontaneously
nucleate magnesium sulfite solids from scrubber type solutions
at various temperatures. It was also important to determine
which hydrate species are nucleated at the various temperatures
and whether any hydrate mixtures are obtained.
312

-------
SECTION 2
EXPERIMENTAL METHOD FOR HOMOGENEOUS NUCLEATION
Concentrated sodium sulfite solutions were added to
aliquots of a concentrated magnesium sulfate solution at vari-
ous temperatures until nucleation was visually observed. The
final compositions of the mixed solutions were comparable to
the magnesium oxide scrubber conditions : 207o magnesium sulfate
(2.08 molal) and pH 6. The temperature range investigated was
28 to 80°C.
EXPERIMENTAL DETAILS
A stock solution of reagent grade magnesium sulfate was
prepared and analyzed to contain 2.63 molal (or 2.53 molar)
magnesium and total sulfate. The pH at room temperature was
2.9. Insoluble traces which otherwise might have interfered
with the nucleation process were allowed to settle to the
bottom of the stock solution for one week, and aliquots were
withdrawn by pipet from the center of the solution.
Nucleation runs were started by transferring lOOg of mag-
nesium sulfate stock solution into a 125-ml erlenmeyer flask
containing a teflon coated magnetic stir bar. The flask was
stoppered and placed into a 0.5 A glass-lined water bath main-
tained at the desired temperature within ±1°C. Solution
stirring, the rate of which was constant from run to run, was
achieved by the use of a magnetic stirrer located under the
water bath. Sodium sulfite salt was added and dissolved in the
magnesium sulfate solution to the saturation limit. Depending
upon the temperature, 0.6 to 2g of sodium sulfite was required.
The pH of the magnesium sulfate solution increased from 2.9 to
approximately 6. A separate sodium sulfite solution was then
313

-------
prepared by dissolving 6g of sodium sulfite in 22g of 0.6 molar
sulfuric acid solution. This solution (pH = 7.5) was saturated
with respect to sodium sulfite at 25°C and was subsaturated at
the higher temperatures.
After temperature stabilization was obtained, the sodium
sulfite solution was added dropwise to the magnesium sulfate
solution at the rate of 4-5 ml per minute. At the first visual
sign of nucleation, the pH was measured with a temperature-
compensated Orion 701 meter and the samples were diluted and
analyzed. The elapsed time from start to finish of each
nucleation experiment was generally less than seven minutes.
After nucleation, the slurry was immediately filtered, and the
solids were washed with 50% ethanol and oven-dried at 45°C for
30 minutes. The oxidation of sulfite ion was negligible be-
cause of the short time required to perform these nucleation
runs. The analytical methods used were:
Magnesium - Colorimetric titration with an EDTA
standard solution.
Sodium Sulfite - Iodine-buffer preparation and
back-titration with a thio-
sulfate standard solution.
Sulfate - Peroxide oxidation of sulfite, hydro-
gen ion exchange, controlled evaporation
leaving quantitative residue of sulfuric
acid, and titration with a hydroxide
standard solution. Sulfate = total sul-
fur minus sulfite.
314

-------
SECTION 3
RESULTS
Primary nucleation runs were made at 28, 40, 50, 69, and
80°C in 207o magnesium sulfate media. These data are shown in
Table 3-1. The magnesium, sodium, sulfite, and sulfate con-
centrations are expressed in molality units.
Identification of the nucleated solids was made by dif-
ferential scanning colorimetry (DSC) using the Perkin Elmer
DSC-2 instrument. The waters of hydration of magnesium sulfite
hexahydrate or trihydrate are released in one step by using a
steady stream of nitrogen across the sample. The loss of water
is recorded in the DSC thermogram by measuring the surge of
energy required to maintain a steady temperature rise. The
DSC thermograms of the various nucleated solids are shown in
Figure 3-1. Two patterns are observed: the loss of the waters
of hydration at 100-110°C or at 160-170°C, conforming, respec-
tively, to the properties of pure MgS03-6H20 or pure MgS03-3H20.
Chemical analyses showed that the magnesium sulfite content
of the nucleated solids was either 4.7 or 6.3 mole/g, which is,
respectively, the theoretical value for MgS03*6H20 or
MgS03'3H20. These results and the DSC thermograms above showed
that pure MgS03*3H20 is nucleated at 69 and 80°C. No hydrate
mixtures were observed.
The activity products
aMg-H- ' asot ' a3H20 * ap3
were calculated by the Radian equilibrium program using the
data from Table 3-1. The ap3 values versus nucleation
315

-------
TABLE 3-1. SOLUTION COMPOSITIONS OBSERVED TO PRODUCE
HOMOGENEOUS NUCLEATION AT VARIOUS TEMPERATURES
Temperature

{Mg+2}


Species

°C±1°C
PH
{Na2S03>
{so*-2}
Nucleated
ap 3xl05
28
5.7
2.220
0.381
2.240
MgSOs•6H20*
30.5
40
6.0
2.190
0.393
2.206
MgS03-6H20**
31.6
50
6.1
2.028
0.519
2.053
MgS03-6H20**
35.9
60
6.2
2.091
0.602
2.112
MgS03-6H20**
37.1
69
6.2
2.040
0.678
2.065
MgS03-3H20
35 .3
80
6.3
2.257
0.602
2.273
MgSO 3•3H20
28.8
Note Solution concentrations are expressed in molality units. Density was 1.260
g/ml at room temperature.
* Rhombic crystal habit
** Hexagonal crystal habit

-------

-------
temperatures are plotted in Figure 3-2, and curves have been
drawn to represent the boundary for the onset of homogeneous
nucleation for the hexahydrate and trihydrate phases of mag-
nesium sulfite. Also included for comparison are the saturated
solution activity products of MgS03-6H20 and MgS03-3H20 as a
function of temperature. At the start of nucleation the rela-
tive saturation of MgS03"6H20, defined as ap6/Ksp6, was approxi-
mately four at 28°C and approximately three at 60°C, where
ap6 = aPl • a3HzQ .
At 69 and 80°C, the relative saturation of MgS03-3H20, defined
as ap6/Ksp3, was approximately twelve when nucleation occurred.
318

-------
40.0
Activity Products In Supar<
saturated Soiuclona Vfh«n
Nuclastlon Occurred
36.0
2a.o
24.0
16.0
12.0
3ATURATC0
SOUmON
ACTIVITY CROOUCT8
a.o
MgSOa•3H,0
4.0
0.0
30
20
0
40
SO
70
ao
so
10
80
TEMPERATURE (#C)
Figure 3r2. Homogeneous Nucleation from 20%
MgSOt, Solutions and pH 6
319

-------
SECTION 4
DISCUSSION
Magnesium sulfite trihydrate is predicted by thermodynamics
to be the stable phase in 207o magnesium sulfate solutions at
AO, 50 and 60°C. However, from Section 3 it was shown that
pure magnesium sulfite hexahydrate was nucleated at these
temperatures. Only at temperatures above 65°C was the tri-
hydrate phase observed to nucleate. This indicates that a
kinetic effect is definitely involved in the nucleation process.
The shape of the hexahydrate crystals nucleated at 28°C
was different from that of hexahydrate crystals obtained at the
higher temperatures. However, the x-ray diffraction patterns
of the solids nucleated at 28 and 50°C were virtually the same
and confirm that these crystals are pure MgS03-6H20 (see Figures
4-1 and 4-2). The two patterns compare well with each other and
with the literature values for MgS03-6H20 listed in Table 4-1.
The relative intensities, I/Ij, of the peaks depend somewhat
on the packing of the crystals in the sample holder but are
mostly a function of the atomic spacings in the crystal lattice.
Positive identification of crystals is possible by this method.
The x-ray diffraction pattern (Figure 4-3) of the solids
nucleated at 60°C was also obtained and it was found to agree
with the literature data for MgSOaOHaO (Table 4-2). This
comparison confirms that magnesium sulfite was nucleated as the
pure trihydrate phase at 69°C.
Microscopic examination revealed the characteristic
shapes and sizes of the various types of nucleated crystals.
Four samples are discussed here:
320

-------
t§
5 °
M
x -
il
o
200O
1SOO
1000
500
AlIaA
A
CO
N>
Figure 4-1. X-Ray Diffraction Pattern of MgS03«6H20
Crystals Nucleated at 28°C
M

zooo
isoo
1000
900
LI

ss
M	SS	SO
TWO 1NCTA AMI CDCOTEES
Figure 4-2. X-Ray Diffraction Pattern of MgS03*6Hz0
Crystals Nucleated at 50°C

-------
TABLE 4-1. X-RAY DIFFRACTION DATA FOR
MgS03«6H20
d(A°)
I/Ii
20 for CuKa
(degrees)
5.7
4
15.5
4.40
80
20.2
3.87
100
23.0
3.01
32
29.6
2.90
32
30.8
2.74
100
32.6
2.42
60
37.1
2.05
12
44.1
1. 91
32
47.6
1.77
70
51.6
1. 71
4
53.6
1. 66
32
55.3
1.63
4
56.4
1. 54
8
60.0
ASTM X-Ray Powder Diffraction Pattern Data File, Card
No. 1-0473.
322

-------
1000
I 800
u
S 800
m
t-
z
400
30
25
35
20
15
40
10
45
55
50
60
N>
OJ	TWO THETA ANGLE (DEGREES)
Figure 4-3. X-Ray Diffraction Pattern of MgS03*3H20
Crystals Nucleated at 69°C

-------
TABLE 4-2. X-RAY DIFFRACTION DATA FOR MgS03«3H20
d(A°)
I/I
20 CuKai
degrees
d (A0)
I/I
20 CuKa
degrees
6. 71
80
13.2
2.23
8
40.4
4.70
30
18.9
2.20
6
41.0
4.27
60
20.8
2.11
6
42.8
3.62
20
24.6
2.08
4
43.5
3. 35
100
26.6
2.02
14
44.8
3.02
8
29.6
1.99
6
45.6
2.98
30
30.0
1.89
4
48.0
2.87
10
31.1
1.84
4
49.5
2.77
6
32.3
1.81
6
50.5
2.72
20
32.9
1.80
14
50.8
2.65
30
33.8
1.78
8
51.3
2.62
45
34.2
1.76
30
51.9
2.56
50
35.0
1.74
8
52.6
2.39
20
37.6
1.69
20
54.3
2. 37
40
37.9
1.67
6
55.0
2.32
40
38.8
1.63
4
56.4



1.56
4
59.0
ASTM X-Ray Powder Diffraction Pattern Data File, Card No. 11-239
324

-------
MMI
0 .

FIGURE 4-4.
PHOTOMICROGRAPH 125X
OF MgSO3»6H20 CRYSTALS (RHOMBIC)
NUCLEATED AT 28°C
325

-------
FIGURE 4-5.
PHOTOMICROGRAPH 125X
MgS03 • 6H20 CRYSTALS (HEXAGONAL)
NUCLEATED AT 50°C
326

-------
FIGURE 4-6.
PHOTOMICROGRAPH 250X
OF MgS03 • 3H20 CRYSTALS NUCLEATED
AND GROWN OVERNIGHT AT 65° C
327

-------
FIGURE 4-7
ELECTRON MICROGRAPH 300X
OF MgS03 . 6H20 CRYSTALS (RHOMBIC)
NUCLEATED AT 28°C
328

-------
FIGURE 4-8.
ELECTRON MICROGRAPH 100X
MgS03*6H20 CRYSTALS (HEXAGONAL)
NUCLEATED AT 50°C
329

-------
FIGURE 4-9.
ELECTRON MICROGRAPH 1000X
OF MgS03 • 3H20 CRYSTALS
NUCLEATED AT 69°C
330

-------
FIGURE 4-10
ELECTRON MICROGRAPH 1000X OF
MgS03 • 3H20 CRYSTALS NUCLEATED
AND GROWN OVERNIGHT AT 65° C
331

-------
(1)	HgS0346H20 nucleated	at 28°C
(2)	MgS03*6H20 nucleated	at 50°C
(3)	MgS03*3H20 nucleated	at 69°C
(4)	MgS03*3H20 nucleated	and grown overnight at 65°C.
Figure 4-4 shows the MgS03»6H20 crystals that were nucle-
ated at 28°C. These crystals are of the "rhombic" habit and
are characterized as pyramid shaped crystals having triangular
faces and sharp points. Individual crystal sizes vary from 10
to 100 microns.
The MgS03*6H20 crystals nucleated at 50^ (Figure 4-5) are
typical of the other hexahydrate crystals obtained by nucleation
in 20% magnesium sulfate solutions at 40 and 60°C. These crys-
tals are of the "hexagonal" habit and are characterized as box-
shaped crystals with raised flat surfaces at the corners. Size
variation was 10-100 microns.
The MgS03'3H20 crystals nucleated at 69°C were visible
only at 400x magnification and no photomicrograph is included
here. Larger trihydrate crystals, which were prepared in a
separate test by nucleation and overnight growth in a 20% mag-
nesium sulfate solution at 65°C, are shown in Figure 4-6. These
crystals ( 10 microns) are barely distinguishable in this photo.
Electron micrographs of the four samples listed above were
also made (Figures 4-7 through 4-10). These film exposures
have the advantage of providing a greater depth of field. Some
deterioration of the crystals such as a loss of waters of hydra-
tion was caused by the electron bombardment and the high vacuum
conditions. Gold plating of the crystal samples was performed
332

-------
in an effort to reduce these effects. Figures 4-7 and 4-8
provide a comparison of the MgS03*6H20 crystals nucleated at
28 and 50°C, respectively. Unfortunately, in the case of the
"hexagonal" crystal habit, some detail is lost by the rounding
of some edges and corners of the box-shaped crystals. The
MgSO3*31120 crystals nucleated at 69°C are shown in Figure 4-9.
Individual microsized crystals are noticeable even though the
details are lacking. The trihydrate crystals nucleated and
grown at 65°C are more well defined. See Figure 4-10. These
crystals are characterized as bipyramid shaped crystals with
size variation of 1-10 microns in length.
The possibility of sulfate inclusion in the nucleated
solids was investigated. In spite of the high magnesium sul-
fate concentrations, no sulfate content in the crystals was
found. Infrared spectra of nucleated magnesium sulfite hexa-
hydrate and trihydrate crystals are shown in Figure 4-11. Sul-
fate, which would absorb at 1100 cm"1, is absent from these
spectra.
333

-------
MgSOg • 6H2O NUCLEATED
AT BO'C
t ¦ -	*	*	¦	1	_—— —	<	,— 1	i
M33O3 • 3H20 NUCLEATED
AT 60*C
BOO
1000
3000
Figure 4-11. Infrared Spectra of Magnesium
Sulfite Nucleated at 50°C and 69°C
334

-------
SECTION 5
CONCLUSION
The magnesium sulfite solids that nucleated from the 20%
magnesium sulfate pH 6 solutions consisted of pure MgS03*6H20
at 28, 40, 50, and 60°C and pure MgS03*3H20 at 69 and 70°C. No
hydrate mixtures were observed. The relative saturation values
required to cause homogeneous nucleation were approximately 3-4
for MgS03»6H20 and approximately 12 for MgS03*3H20. Sizes of
the nucleated crystals ranged from 1-10 microns for the tri-
hydrate phase and 10-100 microns for the hexahydrate phase of
magnesium sulfite.
335

-------
TECHNICAL NOTE 200-045-54-05
SECONDARY NUCLEATION OF MgS03
HYDRATES IN SCRUBBER-LIKE MEDIA
Prepared by:
L. A. Rohlack
R. E. Pyle
336

-------
CONTENJS
1.	Introduction	338
2.	Experimental approach			339
Secondary Nucleation Apparatus 		339
Operational Procedure	346
Analytical Methods 		348
3.	Experimental Results	35O
Data Processing	350
Results	352
4.	Discussion of Results	356
FIGURES
Number	Page
2-1 Secondary Nucleation Apparatus 		340
2-2 Plexi-Glas Reactor 		342
2-3	Continuous pH Monitor	345
3-1	MgS03*6H20 Secondary Nucleation Run #4-3B	351
3-2 Primary and Secondary Nucleation from 20% MgSO*
Solutions and pH 6	355
TABLES
Number	Page
3-1 Secondary Nucleation Experimental Run Data 		353
3-2 Secondary Nucleation Experimental ?vun Data	354
337

-------
SECTION 1
INTRODUCTION
Radian Corporation has conducted an experimental study to
identify, measure, and characterize the formation of magnesium
sulfite trihydrate and hexahydrate solids by the phenomenon of
secondary nucleation. This study is essential in order to
accurately predict conditions necessary for the preferential
production of either hydrated phase of MgS03• It was therefore,
the objective of this study to:
determine the onset of secondary nucleation
for the trihydrate and hexahydrate phases of
MgS03, and
characterize the influence of temperature and
stirring rate on the threshold or onset for
secondary nucleation.
This Technical Note presents in detail the experimental
approach used and the experimental data obtained from the secon-
dary nucleation kinetics studies on the hexahydrate and tri-
hydrate crystalline phases of MgS03.
338

-------
SECTION 2
EXPERIMENTAL APPROACH
An experimental secondary nucleation apparatus utilizing
a continuous liquid feed, batch-solid system was specially de-
signed and constructed to perform the experimental segment of
this study. During one phase of the secondary nucleation of
magnesium sulfite trihydrate experiments, a slurry recycle
loop was incorporated into the kinetics apparatus. Along with
the design of the kinetics apparatus, special consideration
was given to both the operational and analytical procedures
used during this study in order to ensure that consistent and
accurate measurement of all pertinent parameters could be ob-
tained.
Secondary Nucleation Apparatus
The secondary nucleation apparatus used during this study
is a crystallization kinetics device from which quantitative
measurements can be obtained for correlating the variables of
concentration, temperature, and stirrer speed with the secon-
dary nucleation of insoluble solids. With this apparatus,
experimental conditions such as reactor temperature, stirring
speed, feed stream composition and flowrate can be controlled
and varied independently to achieve conditions required for the
onset of secondary nucleation.
A schematic diagram of the apparatus is shown in Figure
2-1. The essential components are:
continuous liquor feed system,
reactor, and
339

-------
	\
Co
O
FEED
NO. 1
ROTOMETER
——dc
r
GZ3
PARASTALTtC PUMP
AND CONTROLLER
ROTOMETER
FEED
NO. 2
PARA8TALT1C
MO CONTROLLER
Figure 2-1.
CONTROLLED
TEMPERATURE
BATH
FEED
SAMPLE
STWRER
-Q_£LQ-
X8JL
SAMPLE
Ob

-------
pH monitor.
Continuous Liquor Feed System
The continuous liquid feed system supplies the reactor
with two well defined feed streams. Parameters such as liquor
composition, flowrate, and temperature can be varied and
accurately controlled. Two 60-liter linear polyethylene con-
tainers were used to mix and store the feed stream solutions.
Peristaltic pumps and controllers were used in transferring
the feed streams from the feed tanks to the reactor at accu-
rately controlled flowrates which were easily monitored by an
inline rotameter. Before entering the reactor, the feed
streams passed through a constant temperature water bath so
that the reactor temperature could be varied as desired from
30°C to 68°C. When reactor temperatures in excess of 40°C were
desired, the reactor was placed in the temperature-controlled
bath to minimize heat loss to the atmosphere. Upon leaving
the constant temperature bath the feed streams passed through
a three-way stopcock. At this point the feed streams could
either go to the reactor or could be diverted to the feed
stream sampling points.
Plexiglas Reactor
A specially designed 2750 ml plexiglas reactor, as shown in
Figure 2-2, served as the reaction vessel for precipitation and
incipient secondary nucleation. The feed streams were input
to the reactor at an angle of 180 degrees to each other and
directed toward a propeller to insure rapid and thorough mixing
of the feed streams. A second propeller near the bottom of
the reactor was used to suspend the solids while baffles broke
up the circular motion of the liquor to enhance mixing. Oxygen
free nitrogen gas could be forced into the top of the reactor
to either purge the system of oxygen or force liquor out of the
341

-------
STIRRER
SUPPORT
STIRRER
j THERMOMETER
M A, <1.
Figure 2-2. Plexi-Glas Reactor
342

-------
reactor. The reactor liquor temperature was monitored by a
thermometer mounted in the top of the reactor. In order to sep-
arate the liquor from the solids, a 142 millimeter filter was
positioned in the bottom of the reactor. Liquor passing through
the filter exited the reactor through a plexiglas tube where it
could then be discarded or collected for analysis or storage.
In an attempt to characterize the secondary nucleation of
magnesium sulfite trihydrate, a slurry recycle loop was incor-
porated into the reactor. Slurry was removed from the bottom
of the reactor through a plexiglas tube by means of a peristal-
tic pump. Upon leaving the pump, the slurry was forced through
a quick disconnect and then returned to the top portion of the
reactor. By incorporating this slurry recycle device, a repre-
sentative weight percent solids sample could be obtained quickly
and easily by collecting a known small volume of slurry at the
quick disconnect.
pH Monitor
As the effluent stream left the reactor, its pH was moni-
tored either continuously or intermittently depending upon the
characteristics of the liquor. During preliminary magnesium
sulfite hexahydrate (MgS03'6H20) secondary nucleation runs, in
which the reactor temperature was less than 40° Centigrade, a
continuous pH monitoring system was used to closely monitor the
pH of the effluent stream. This type of pH monitoring system
proved to be an efficient means of obtaining accurate effluent
pH values at low sample temperatures.
During runs made at temperatures greater than 40° Centi-
grade an intermittent pH sampling system was used. The inter-
mittent system proved suitable for obtaining accurate pH values
at higher reaction temperatures.
343

-------
Continuous pH Monitor
The continuous pH monitor, Figure 2-3, was designed to
provide a means for obtaining accurate pH values with minimal
operator effort. Effluent liquor entered the top of a 100
milliliter three-necked flask by means of a glass tube which
extended down into the flask to just above the stirrer. A
small magnetic stirring bar was used to thoroughly mix the
liquor assuring a homogeneous sample was being monitored. pH
electrodes were tightly fitted in both of the side ports of
the flask and the electrodes extended below the surface of
the liquor, insuring proper liquor to electrode contact. Li-
quor left the monitor by means of a glass tube which extended
only slightly down into the flask. The pH meter could be left
on for continuous monitoring of the effluent or the pH meter
could be turned on and off for occasional monitoring of the pH.
Intermittent pH Monitor
Whenever the reactor temperature was less than 40° Centi-
grade, the above pH monitoring system was used because the
reactor temperature and pH monitor sample temperature were
essentially the same. When reactor temperatures above 40°C
were required, the continuous pH monitor proved to be an un-
suitable means for sample pH measurement because of temperature
differences from reactor to sample. Therefore, an inter-
mittent pH monitor was used to give accurate pH measurements of
reactor solutions at temperatures above 40° Centigrade, A
small 50 milliliter beaker served as the collection device.
In the bottom of the beaker a small stirring bar maintained a
homogeneous solution while two pH electrodes submerged in the
liquor monitored the pH. Because of the small size of the
collection device, the time required to obtain an accurate pH
value was minimal, allowing little or no temperature change in
the effluent temperature during pH monitoring. During a run,
pH values were generally obtained during the collection of
344

-------
REACTOR EFFLUENT
IN
PH
ELECTRODE
W
¦*>
Ui
J
100ml FLASK


OFF
7/\Y
?
OFF 11 J J
\j
ON


STIRRER
Figure 2-3. Continuous pH Monitor

-------
aqueous sulfite samples.
OPERATIONAL PROCEDURE
In this section the operational procedures used in perform-
ing the secondary nucleation experiments for both the hexahydrate
and trihydrate forms of magnesium sulfite are discussed. All
of the operational procedures used during this segment of the
program were developed to minimize undesirable changes in either
the solid or liquid phases within the feed tanks, transfer lines,
or reactor. Steps were also taken to ensure maximum ease of
operation during the kinetics run to enable the operator to
concentrate on the progress of the experiment. For a given set
of operating conditions, the following operational procedures
were developed to fulfill the objectives of the run.
Initially, the temperature bath was turned on and the proper
temperature setting selected. Then the two feed tanks were
thoroughly cleaned and filled with the appropriate quantity of
deionized water. In order to minimize oxidation during a run,
oxygen free nitrogen gas was bubbled through the feed tanks, thus
purging any dissolved oxygen from the deionized water and forming
a blanket of inert nitrogen gas above the water. This protective
blanket of nitrogen gas also minimized the chances of air to
liquid contact when the liquid level decreased during the run.
During the purging with nitrogen gas, there was sufficient
time to thoroughly clean the reactor. All foreign material was
removed by first rinsing the reactor with a dilute HC1 solution
which helped loosen and dissolve any scale which may have formed
during the previous run. After sufficient HC1 solution had been
rinsed through the reactor, a sufficient amount of deionized
water was used to finish cleaning out the reactor. After rinsing,
the reactor was examined for physical abnormalities which were
346

-------
corrected before final assembly of the unit. Oxygen-free nitrogen
gas was used to purge any oxygen in the reactor prior to the run.
After final inspection of the reactor, the proper quantities of
chemicals were added to the already purged feed tanks. Then,
depending on the amount of agitation required to completely dis-
solve the chemicals in the feed tanks, either manual stirring or
a laboratory stirrer was used to agitate the liquor. A flotation
lid was used whenever possible to minimize liquid-air contact and
a plexiglas lid was kept on top of the feed tanks at all times
in order to keep out foreign particles and also minimize any air
turbulence above the flotation lid.
After the feed tanks were properly prepared and the reactor
was ready for use, feed stream flow rates were set and feed
stream samples were taken. The peristaltic pumps used to control
the flowrates of the two feed streams were turned on and flow
was diverted to the sample ports. Rotameters gave an accurate
indication of the flowrate and were used to maintain a constant
feed stream flow throughout the course of the run. By adjusting
the peristaltic pump controller, the flowrate could be adjusted
and controlled. The rotameters were calibrated routinely prior
to each run by collecting a certain volume of sample in a speci-
fied time. After a certain feed stream flowrate was obtained,
feed stream samples were collected in the proper collection
medium for storage or analysis.
Next, the magnesium sulfate flow was diverted back into the
reactor and the reactor was allowed to fill approximately 80%
full with the magnesium sulfate feed. The reactor stirrer was
turned on and a stirring rate set by adjusting the variac. For
the majority of the runs the stirring rate was maximum (^1200
RPM), allowing for optimum agitation of reactor contents. A
stirring speed of %1000 RPM (variac set at 85) was used during
347

-------
those runs designed to study the effect of stirring speed on
secondary nucleation. Next, a solution containing approximately
0.5 molar sodium sulfite was added to the reactor in order to
obtain an initial reactor sulfite concentration of .1 molar.
Just before the reactor was completely filled, a known quantity
of seed crystals was added to the agitated liquor. For most of
the runs one gram of seed crystals was used, but during some of
the magnesium sulfite trihydrate runs 10 grams of seed crystals
were used in order to help clarify secondary nucleation of the
trihydrate phase. After the seed crystals were added, the ther-
mometer was positioned. As soon as the reactor was filled with
the .5 molar sulfite solution, the stirrer fitting was tightened
and the feed streams were diverted into the reactor. At this
time the run was officially started.
During all of the runs, several operational variables were
monitored in order to maintain a constant set of working condi-
tions. Feed stream flowrates were kept at a constant rate by
a visual check of the flowrate as indicated by the rotameter.
A thermometer in the top of the reactor allowed for the continu-
ous monitoring of reactor temperature. Adjustments in reactor
temperature could be made by adjusting the reactor bath tempera-
ture, but after several runs the temperature bath setting re-
quired for a certain reactor temperature was accurately known.
For the most part, adjustments were seldom required in either
the feed stream flowrate or the reactor temperature during a run.
This ease of operation allowed the operator to more closely
monitor the approach and occurrence of secondary nucleation.
ANALYTICAL METHODS
In this section, a brief outline of the analytical methods
used in the MgS03 hydrate secondary nucleation studies is given.
348

-------
Magnesium - colorimetric titration with
NaaEDTA or dilution preparation and
atomic absorption determination;
Sulfite - iodine-buffer preparation and
back titration with standard arsenite
or thiosulfate;
Sulfate - peroxide oxidation of sulfite,
hydrogen ion exchange, and titration
with standard sodium hydroxide. Sul-
fate is calculated as the difference
between total sulfur and sulfite
sulfur;
Sodium - dilution preparation and
atomic absorption determination; and
Chloride - potentiometric titration
with standard silver nitrate.
349

-------
SECTION 3
EXPERIMENTAL RESULTS
DATA PROCESSING
The onset of secondary nucleation of magnesium sulfite
hexahydrate is relatively easy to detect by monitoring the
effluent sulfite concentration. Also, because of the sub-
stantial increase in the particle number density that accompanies
secondary nucleation, a visual change in reactor opacity indi-
cates to the operator that secondary nucleation of the hexa-
hydrate crystalline phase has occurred. Because of the substan-
tially slower growth rate of the magnesium sulfite trihydrate
phase the detection of secondary nucleation of this phase by
either monitoring the sulfite concentration or by visual means
was not possible.
Detection of Secondary Nucleation of Magnesium Sulfite
Hexahydrate
By closely monitoring the effluent sulfite concentration
during an experimental run, one can determine accurately the
onset of secondary nucleation of magnesium sulfite hexahydrate.
By plotting the effluent sulfite concentration versus time one
can accurately determine the sulfite concentration and thus the
complete reactor solution composition at the onset of secondary
nucleation.
Figure 3-1 shows a typical plot of effluent sulfite con-
centration versus time for a MgS03 hexahydrate secondary nuclea-
tion run. The onset of a secondary nucleation is taken at the
maximum experimental sulfite concentration.
350

-------
u>
Ul
0.20
0.18
0.18
O
^ 0.14
z
O
P
<
tt
0.12
O
z
o
u
S! o.io
3
n
0.08
0.06
6 10 15 20 25 30 35 40 46 50 55 60 65 70 75 80 85 90 95
TIME (MM.)
Figure 3-1. MgS03-6H20 Secondary Nucleation Run # 4-3B

-------
Detection of Secondary Nucleation of Magnesium Sulfite
Trihydrate
Because of the substantially slower growth rate of the
magnesium sulfite trihydrate phase, the detection of the onset
of secondary nucleation by this batch solid-flow through liquid
experimental technique was not possible. Other experimental
techniques were tried in an effort to characterize trihydrate
secondary nucleation as a function of temperature. A variable
temperature batch-solid - batch-liquid experimental method was
tried, as well as experiments using the batch-solid flow through
liquid crystallizer with added slurry recycle loop. Both tech-
niques rely on a change in the system weight percent solids as
a function of time or temperature as a means of detecting the
onset of secondary nucleation of trihydrate. These techniques,
however, also proved unsuccessful.
RESULTS
Experimental results for the secondary nucleation kinetics
studies on MgS03*6H20 have been summarized in Tables 3-1 and
3-2. The activity products were calculated by inputting the
pertinent reaction solution information such as concentrations
of magnesium, sodium, sulfite, and sulfate; pH and temperature
into the Radian chemical equilibrium computer program.
In Figure 3-2, a plot of y, where y = a*P"6/a3H20' is
shown versus temperature. The secondary nucleation data for
hexahydrate is plotted along with the primary nucleation data
obtained previously for MgS03 hexahydrate and trihydrate (see
Technical Note #200-045-54-04). The results of reduced stirrer
speed hexahydrate kinetics runs are also plotted in Figure 3-2.
352

-------
TABLE 3-1. SECONDARY NUCLEATION EXPERIMENTAL RUN DATA*
Run
Type Seed
Sti rri ng
Speed
(RPM)
Temp.
(°C)
pH
[Mg+2]
[Na+]
[SO"*2 ]
t • 1
CM
1
O
l/>
4-3B
MgSO 3 *6H 2O
1200
30.0
6.03
2.07
.390
2.10
.195
4-4B
NgS03*6H20
1200
44.4
6.03
2.11
.690
2.14
.345
4-7B
MgS03-6H20
1200
46.0
6.0
2.04
.678
2.07
.339
4-8B
MgS03«6H20
1200
45.0
5.96
1.97
.704
2.00
.352
4-9B
MgS03-6H20
1200
33.6
5.75
1.96
.554
1.99
.277
4-1 OB
MgS03>6H20
1000
32.6
6.10
1.93
.514
1.96
.257
4-1 IB
MgS03*6H20
1000
32.5
6.13
1.97
.526
2.00
.263
4-12B
MgS03-6H20
1200
60.5
6.25
1.91
1.136
1.94
.568
4-12B
MgS03*6H20
1200
60.5
6.25
1.91
1.060
1.94
.530
4-12B
MgSO 3 *6H 20
1200
60.5
6.25
2.07
1.154
2.10
.577
4-12B
MgS03*6H20
1200
60.5
6.25
2.07
1.076
2.10
.538
4-13B
MgS03*6H20
1200
59.2
6.27
2.02
1.038
2.05
.519
4-14B
MgS03-6H20
1000
58.3
6.28
1.96
1.062
1.99
.531
* All analyses are reported in millimoles/liter.

-------
TABLE 3-2. SECONDARY NUCLEATION EXPERIMENTAL RUN DATA*
Run No.
a
Mg+2
a
S03"2
a
H20
K
spG
aPe
Rel. Sat.e
y
4-3B
3.349-01
7.276-04
0.939
6.124-05
1.670-04
2.727
2.019-04
4-4B
3.182-01
1.012-03
0.9312
7.62-05
2.100-04
2.756
2.600-04
4-7B
2.955-01
1.322-03
0.9339
7.86-05
2.59-04
3.295
3.18-04
4-8B
2.900-01
1.021-03
0.9345
7.735-05
1.97-04
2.547
2.417-04
4-9B
3.034-01
7.898-04
0.9380
6.45-05
1.632-04
2.530
1.978-04
4-1 OB
2.935-01
1.053-03
0.9401
6.35-05
2.13-04
3.354
2.568-04
4-11B
3.006-01
1.087-03
0.9388
6.34-05
2.24-04
3.533
2.702-04
4-12B
2.498-01
1.754-03
0.9263
9.78-05
2.77-04
2.83
3.489-04
4-12B
2.531-01
1.612-03
0.9281
9.78-05
2.61-04
2.669
3.262-04
4-12B
2.798-01
1.657-03
0.9215
9.78-05
2.838-04
2.902
3.627-04
4-12B
2.833-01
1.521-03
0.9234
9.78-05
2.67-04
2.730
3.393-04
4-13B
2.749-01
1.544-03
0.9257
9.60-05
2.67-04
2.781
3.368-04
4-14B
2.621-01
1.670-03
0.9268
9.48-05
2.77-04
2.922
3.484-04
* All analyses are reported in milli'moles/liter.

-------
44.0-1
o PRIMARY NUCLCATtON fXWIIMCNTAL 0A7A
StCONDARY MUCLIATION IXMMMiNTAL DATA
(Ms«Or«H20>
MCONOARV NUCLIATIOM CXPtMMfNTAl OATA
IMgCOg'tHjO) RIOUCIO STIRRIM S»IIO
40.0-
36.0
U"l
o
el
28.0-
Pl
24.0
r-i
o

20.0'
3S>
16.0-
12.0-
PURE SOLUTION
ACTIVITY PRODUCTS
8.0-
4.0
0.0
0
30
40
50
10
20
80
80
«0
70
TEMPERATURS (*C)
Figure 3-2. Primary and Secondary Nucleation
From 207, MgSOV Solutions and pH 6
02-aaas-i
355

-------
SECTION 4
DISCUSSION OF RESULTS
As can be seen from Figure 3-2, a stirring speed effect
was observed in the hexahydrate secondary nucleation kinetics
results. The slower stirring speed means that a lower impact
energy is imparted to the crystals, and thus a higher solution
supersaturation (i.e., larger activity product) is required to
reach the onset of secondary nucleation. In the limit of zero
stirring speed, the onset of secondary nucleation will be coin-
cident with the onset of primary nucleation.
356

-------
TECHNICAL NOTE 200-045-54-06
ADDITIVE EXPERIMENTS ON TRIHYDRATE
KINETICS
Prepared by:
J. L. Skloss
357

-------
CONTENTS
1.	Introduction	359
2.	Experimental	360
3.	Experimental Results	361
Bibliography 	 367
TABLES
Number	Page
1 Effect of 50 ppm Additive of the Nucleation and
Precipitation of MgS03*3H20 at 55°C and pH 8.5 . . 363
2	Effect of 50 ppm Additive on the Nucleation and
Precipitation of MgS03*3H20 in High Sulfite
Solutions at 55°C and pH 6.1	364
3	Effect of 50 ppm Additive on the Precipitation of
MgS03*3H20 at 55°C and pH 8.2 Using Seed	365
4	Effect of 50 ppm Additive on the Precipitation of
MgS03-3H20 in 20L MgSO^ Solution at 55°C and pH
6.1 Using Seed	366
358

-------
SECTION 1
INTRODUCTION
The removal of sulfur dioxide from flue gases involves
the precipitation of magnesium sulfite hydrate solids in the
magnesium oxide system. Typical scrubber solutions are super-
saturated with respect to magnesium sulfite at the normal
operating temperature of 55°C. MgS03*3H20 is the stable phase
under these conditions, and any enhancement of the normally
slow precipitation rate of MgS03*3H20 would be an advantage to
the scrubber operation.
One method of potentially modifying the precipitation
rate is the use of metal cations of high ionic charge as addi-
tives (SH-021). The purpose of this study is to investigate
the effect of these various additives on the precipitation and
nucleation kinetics of MgS03*3H20. This Technical Note reports
the experimental results of this study.
359

-------
SECTION 2
EXPERIMENTAL
Separate 55°C solutions of magnesium chloride or magnesium
sulfate and sodium sulfite were prepared and were mixed
together. 500-ml portions were transferred into erlenmeyer
flasks, and the flasks were immersed in a common 55°C water
bath. Magnesium, sulfite, and pH measurements were made. To
study additive effects on precipitation rate, 0.16g of pure
MgS03»3H20 seed was added to each flask. Then 50 ppm of various
additives were added to the test flasks, and the effects were
compared with the control flask which contained no additive.
Each flask contained a magnetic stir bar which was powered by
a separate magnetic stirrer located below the water bath. The
flasks were sealed with rubber stoppers, and the solutions were
stirred continuously for 48 hours. The solutions were super-
saturated with respect to magnesium sulfite at 55°C, and
MgS03*3H20 solids were nucleated or precipitated during this
period. The slurries were filtered, and the solids were
washed, dried, and weighed.
360

-------
SECTION 3
EXPERIMENTAL RESULTS
The results of the additive experiments are presented in
Tables 1-4.
Table 1
The precipitated solids were slightly colored by the
additive agents, but the chemical analyses showed that the
solids were nearly pure MgS03*3H20. No substantial change of
the additive in the solution was found. Urea addition had no
effect.
Table 2
PH lowered to 6.1.
Table 3
Additive agents had no effect, except for Co which caused
some of the seed crystals to dissolve. The Co complex with
_ 2
SO3 is probable.
Table 4
Scrubber condition of 20% MgSOi, and pH 6. Additives had no
effect except for Co which caused a negative effect on the pre-
cipitation of MgS09*3H20.
It is observed that all of the additives except urea appear
to enhance the rate of growth of MgS0s'3H2 0 in those experiments
361

-------
without the addition of seed material; that is, nucleation. The
additives appear to have 110 enhancement effect on the precipitation
rate of MgS03*3H20.
362

-------
TABLE 1. EFFECT OF 50 PPM ADDITIVE OF THE NUCLEATION AND
PRECIPITATION OF MgS03*3H20 AT 55°C AND pH 8.5.
Initial Mg
mole/£
Initial S03
raole/Jl
50 pptn
Additive used
MgS03-3H20
ppt'd, g/SL
0.134
0.135
Al+3 as AlCla
4.2
0.147
0.124
Cr+6 as Cr03
2.0
0.145
0.132
+ 3
Cr as Cr2(S0it)3
1.3
0.147
0.133
Co as C0CI2
3.2
0.145
0.137
urea
0.10
0.135
0.137
none
0.10
Rel.. Sat. = a.p.3/K = 1.5
sp3
363

-------
TABLE 2. EFFECT OF 50 PPM ADDITIVE ON THE NUCLEATION AND
PRECIPITATION OF MgS03-3H20 IN HIGH SULFITE
SOLUTIONS AT 55°C AND pH 6.1.
50 ppm
Additive used
MgS03*3H20
ppt'd, g/l
A1 as A1C13
0.2
Cr+6 as Cr03
1.0
n
Cr as Cr2(SOi»)3
0.4
Co as C0CI2
0.2
none
0.05
Initial Mg
mole/£
Initial SO3
mole/Jl
0.174
1.48
0.174
1.48
0.174
1.48
0.174
1.48
0.174
1.48
Rel. Sat. = 2.3
364

-------
TABLE 3. EFFECT OF 50 PPM ADDITIVE ON THE PRECIPITATION
OF MgS03«3H20 AT 55°C AND pH 8.2 USING SEED.
Initial Mg
mole/H
Initial SO3
mole/%
50 ppm
Additive used
MgS03'3H20
ppt'd, g/&
0.139
0.1 AO
0.132
0.1 J7
0.132
0 , .11\ 1
0.130
0.135
0.132
0.137
0.132
0.132
Al+3 as AICI3
Cr+6 as Cr03
Cr+3 as Cr2(SOi»)3
| |
Co as C0CI2
urea
none
0.76
0.77
0.81
0.69
0.76
Rel. Sat =» 1.5
Amount of seed used was 0.32 g MgS03*3H20 per Si. g/S, MgS03*3H20 preci-
pitated are in excess of the seed.
365

-------
TABLE 4. EFFECT OF 50 PPM ADDITIVE ON THE PRECIPITATION
OF MgS03'3H20 IN 20% MgSO^ SOLUTION AT 55°C
AND pH 6.1 USING SEED.
Initial Mg
moIe/£
2.2
Initial SO3
mole/Jt
0.30
50 ppm
Additive used
Cr+6 as K2Cr207
MgS03*3H20
ppt'd, k/1
17.8
2.2
2.2
2.2
2.2
2.2
0.30
0.30
0.30
0.30
0.30
Co as C0CI2
Ni44" as NiCl2
j |
Cu as CuCl2
Zn44 as SiSOi*
none
4.1
22.8
19.1
21.8
22.2
Amount of seed used was 0.32g MgS03-3H20 per Z. g/£ MgS03*3H20
precipitated are in excess of the seed.
366

-------
BIBLIOGRAPHY
SH-021 Shor, Steven M. , Effects of Surfactants and Inorganic
Additives on Nucleation Kinetics in Mixed Suspension
Crystallization. PhD. Thesis, Iowa State University,
1970.
367

-------
TECHNICAL REPORT DATA
(Please read InUructions on the reverse be/ore completing)
\ . RLPOR 1 NO. 2.
EPA-600/7-77-109
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Precipitation Chemistry of Magnesium Sulfite
Hydrates in Magnesium Oxide Scrubbing
5.	REPORT DATE
6.	PERFORMING ORGANI7A I ION CODE
7. AUTHOR(S)
Philips. Lowell, Frank B. Meserole, and
Terry B. Parsons
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
8500 Shoal Creek Boulevard
Austin, Texas 78766
10.	PROGRAM ELEMENT NO.
EHE528
11.	CONTRACT/GRANT NO.
68-02-1319, Tasks 36 and 54
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13.	TYPE OF REPORT AND PERIOD COVERED
Task Final; 7/75-12/76
14.	SPONSORING AGENCY CODE
EPA/600/13
15. suppleme ntary notes jerl-RTP Task Officer for this report is Charles J. Chatlynne,
Mail Drop 61, 919/541-2915.
16. abstract The report gives results of laboratory studies defining the precipitation
chemistry of MgS03 hydrates. The results apply to the design of Mg-based scrubbing
processes for S02 removal from combustion flue gas. In Mg-based scrubbing pro*
cesses, MgS03 precipitates as either trihydrate or hexahydrate. The hydrate formed
depends on equipment design and operating conditions. Theoretical prediction, veri-
fied experimentally, indicated that MgS03 trihydrate is the theromodynamically
stable hydrate formed at scrubbing process conditions. MgS03 hexahydrate is formed
as a metastable solid due to kinetic phenomena. Nucleation and crystal growth rates
are much faster for hexahydrate than for trihydrate. The time scales observed in
kinetic experiments at scrubbing process conditions are: hexahydrate precipitation
(10's of minutes), hexahydrate dissolution and trihydrate precipitation (100's of min-
utes), and attainment of trihydrate equilibrium (1000's of minutes). Nucleation plays
a dominant role in the formation of trihydrate solids. These results indicate that
Mg-based scrubbing process can be designed to precipitate a majority of either
hydrate form. Important design variables include scrubbing liquor composition and
temperature, seed crystal composition, slurry volume, equipment residence times,
and energy inputs to the slurry that influence nucleation.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. cor. at 1 I'ieltl/C.nmji
13B
07B 07D
07A
2 IB
Air Pollution Precipitation (Che-
Magnesium Oxides mistry)
Scrubbers Desulfurization
Magnesium Inorganic Flue Gases
Compounds Combustion
Hydrates Nucleation
Air Pollution Control
Stationary Sources
Magnesium Sulfite
Hydrates
18. UIS 1 HIUU riON ST ATE ML-NT
Unlimited
19. GECUHil V CLASS (t his fU port)
Unclassified
21. NO. U( rAL.lt>
374
70. SECURI1 V CLASS (This [>nge.)
Unclassified
22. I'RICI.
EPA Form 2220-t (9-73)
368

-------