Addressing “Impossible to Measure”
pH Applications in the Inorganic
Pigments Industry
A white paper by Endress+Hauser, Inc.
Eric Pfannenstiel, author
Eric Pfannenstiel is a Business
Development Manager with
Endress+Hauser, Inc. in
Greenwood, Indiana. Previously
he served as a Principle Account
Manager with Foxboro. He has
more than twelve years in
laborator y
and
process
analytical instrumentation. Eric
obtained his Bachelor of Science
in Chemistry from Adams State
College in Alamosa, Colorado.
© 2000 Endress+Hauser, Inc.
Intr
oduction
Introduction
The pigments industry is characterized by the manufacture of a
number of inorganic and organic substances that are produced
and marketed as fine powders. These products are used as
decorative or protective coatings and colorants for plastics, fibers,
paper, rubber, glass, cement, glazes, porcelain enamels, printing
inks, and even cosmetics. Inorganic pigments are simple materials
which include basic elements — oxides, mixed oxides, sulfides,
chromates, silicates, phosphates, and carbonates. They are
generally classified by color or physical properties. The table below
shows common pH pigments.
White Pigments
Colored Pigments
Titanium dioxide
Zinc oxide
Zinc Sulfate
Lithopones
Lead Whites
Other Whites
Iron Oxides
Complex Inorganic
Pigments
Mixed metal Oxides
(Spindels, Rutiles,
Zircon Pigments)
Ultramarine Pigments
Cyanide
Iron Blues
Cadmium Pigments
Lead Chromate
Pigments
Black
Pigments
Extenders &
Opacifiers
Carbon
Blacks
Miscellaneous
Pigments
Luminescent Pigments
Metal Effect Pigments
Nacreous Pigments
Transparent Pigments
iThe commercial manufacturing of these materials is very complex
and demands rigorous attention to ever y aspect of the
manufacturing process including tightly controlled pH measurement.
Proper pH control throughout the process is one of several variables
that contribute to final product quality as determined by both
physical and chemical properties. Important physical characteristics
include particle size and distribution, particle shape, and
agglomeration. Chemical properties include chemical composition,
crystalline structure, product purity, and material stability. Control
of each property is essential to insure uniform color dispersion and
opaqueness throughout the materials to which they are added.
The particle size and the difference, between the refractive index
of the pigment, and that of the dispersed media determine pigment
opacity. A particle having a size of .16 – .28 mm provides maximum
visible light dispersion. For inorganic pigments to be useful in most
applications, average particle size must be between .1 and 10 mm.
Any agglomeration of pigment particles can affect its opacity. Thus
insuring optimal particle size, distribution, and preventing
agglomeration is essential to achieve maximum pigment opacity.
One means of accomplishing this is by carefully varying the process
pH and temperature to achieve desired particle size. Other key
physical properties include lightfastness, ability to resist weathering,
heat stability, and chemical resistance.
1
Titanium dioxide and iron oxides are the two most prevalent
inorganic pigments manufactured globally. Annual world demand
is approximately 3,555 x 106 metric tons and 900,000 metric tons
respectively. Carbon blacks are also commonly used. Increasing
environmental concerns have drastically reduced the production
of chromium, cadmium , and lead based pigments. For the purpose
of brevity, this paper describes the manufacturing processes and
application of pH measurement in the production of titanium dioxide
and iron oxide pigments.
Titanium Dioxide Pr
oduction
Production
The base raw material for the production of titanium dioxide is
ilmenite, synthetic rutile or titanium slag. Two forms of titanium
dioxide are produced: anatase and rutile. Anatase is generally less
coarse and is preferred where abrasion is a concern, such as with
wear in thread guides or spinning equipment. Rutile has a higher
refractive index and corresponding opacity and is more commonly
used. Two commercial manufacturing processes, sulfate and
chloride, are used to produce these products.
The sulfate method uses concentrated sulfuric acid to decompose
ilmenite in a digest reactor over approximately a 12-hour period.
The exothermic reaction: FeTiO3 + 2 H2SO4 ®TiSO4 + FeSO4 + 2 H2O
yields approximately 95-97% solubilized TiO 2. Scrap iron is then
added to reduce residual Fe3+ in solution to Fe2+ to allow iron
removal through precipitation. The resulting cake is extracted with
water at a temperature of 65°C, the temperature of maximum iron
sulfate solubility, and the titanium extract filtered off. By cooling the
filtrate solution to 15°C iron (II) sulfate is precipitated in a vacuum
crystallizer. Centrifugation or filtration is used to separate the iron
sulfate from the TiO 2 filtrate. Temperatures throughout these steps
should not exceed 70°C in order to prevent premature hydrolysis of
titanium dioxide.
Hydrolysis of the titanium dioxide mother liquor is initiated by adding
crystallizing seeds to the filtrate at temperatures close to its boiling point
(109°C). The resulting reaction: TiOSO4 + (n+1)H20 ®TiO2•nH2O + H2SO4
must be carefully controlled to insure optimal final product physical
characteristics. To produce anatase titanium dioxide pigment,
anatase microcrystalline seeds are added to the mother liquor in a
concentration of .5 –1% and the mixture hydrolyzed for 3-6 hours.
To produce rutile titanium dioxide pigment, a hydrosol made from a
monohydric acid such as hydrochloric acid is added to neutralize
the mother liquor. This reaction only takes about one hour. In both
instances, free sulfuric acid still entrained in the mother liquor, must
be separated from the resulting hydrolysate to prevent possible
dissolution. Repeated water washing and filtration of the gel carried
out in a vacuum filter removes most of the sulfuric acid. Any residual
is removed during the final calcination process. At this time small
quantities of various chemicals are doped into the solution to
improve the final pigmentary properties.
2
Calcination of the hydrated TiO 2 Gel cake is performed in rotary
kilns with an excess of air to prevent possible reduction of the
titanium dioxide. Water is removed at temperatures between
200-300°C, sulfur trioxide between 480-800°C, and crystals of TiO2
grown at higher temperatures. Final temperature for an anatase
pigment should reach 800-850°C. Rutile white pigment is produced
at temperatures of 900-930°C. This temperature is critical in order
to produce pigment particle sizes of 200 – 400µm. Higher
temperatures produce larger particle sizes that do not exhibit good
pigmentary properties.
The chloride process for manufacturing titanium dioxide accounts
for 56% of the world capacity. Finely ground rutile reacts with chlorine
and calcined coke in a fluidized bed reactor at temperatures
between 800-1200°C: TiO 2 + 2Cl 2 + C®TiCl 4 +CO 2. Oxygen
is added to the reaction in order to maintain the reactor temperature.
All base material added to the reactor is dry to prevent the formation
of hydrochloric acid. Metallic impurities (magnesium, calcium and
zircon) present in the raw titanium feedstock react with chlorine to
form metallic chlorides and accumulate in the bottom of the reactor.
Volatile chlorides including TiCl4 are vented as gases through the
top of the furnace where they are cooled to less than 300°C. Most
impurities, with the exception of vanadium and silicon chlorides,
are separated from the titanium tetrachloride in this step. Vanadium
and silicon chlorides are then removed by reduction to lower chloride
oxidation states and fractional distillation. High purity TiCl4 is
preheated, mixed with hot oxygen, and combusted at 900-1400°C
to form titanium dioxide: TiCl 4 + O2®TiO 2 + Cl2. Aluminum chloride
is added to insure that the final product is rutile. Throughout the
process, factors such as reaction temperature, oxygen levels, water,
and mixing conditions influence the final quality of the titanium
dioxide product.
Titanium dioxide derived from the chlorine process is generally
preferred because it is lighter in color, has lower capital investment
costs, and has less environmental concerns. The disadvantage is
the higher quality feedstocks required and the increased
abrasiveness of the material.
The final step in both processes is pigment finishing. Both rutile
and anatase pigments are coated with inorganic oxides to optimize
dispersability, dispersion stability, opacity, gloss, and durability. The
actual finishing process and coating used is specific to the final
product application and the market sector in which it is used. Rutile
pigments usually receive a 1-15% inorganic coating; anatase a
1-5% coating. The initial finishing stage disperses the base pigment
in water with phosphate, silicate, or organic dispersants. The
suspension is milled and classified to remove oversize particles.
Selective precipitation with small quantities of colorless hydrous
oxides such as P2O5, SiO2, Al2O3, TiO2, and ZrO2 coats the dispersed
particles through specific changes in pH and temperature. Once
coated, the pigment is filtered, washed, and dried. During
subsequent milling, an organic surface treatment such as polyol or
alknolamine is applied for use in paints. Siloxane is added for
plastics. The end product is then filtered, dried, and packaged.
3
Ir
on Oxide Pr
oduction
Iron
Production
Iron Oxide represents approximately 40% of the total production of
colored, inorganic pigments. Yellow geoethite [a-FeO(OH)], orange
lepidocrocite [g-FeO(OH)], red hematite [a-Fe 2O 3], and brown
maghemite [g-Fe2O3] comprise the primary pigment colors. Iron
pigments are characterized by low chroma and excellent
lightfastness. They are nontoxic, non-bleeding, and inexpensive.
They do not react with weak acids and alkalis, they are not
contaminated with manganese, and do not react with organic
solvents. Although naturally occurring deposits are common, most
are not rich enough to allow use in pigmentary applications.
Synthetic iron oxide pigments are preferred for coloring plastics,
paper, rubber, and magnetic recording tapes. Their advantages
include chemical purity, uniform particle size and distribution, and
decreased environmental concerns as compared to their heavy
metal counterparts.
Iron oxide reds are available in colors ranging from orange through
pure red and violet. Varying shades are controlled by particle size,
shape, and surface properties. Four primary methods are used in
the production of iron oxide reds (red hematite [a-Fe2O 3]):
1. Two-stage calcination of FeSO 4•7H20
2. Precipitation from an aqueous solution
3. Thermal dehydration of yellow geoethite
4. Oxidation of black oxide (Fe3O4).
The final product of each process is Fe2O3, however specific physical
properties are determined by the manner of preparation. Thermal
dehydration of yellow goethite yields a product with the lowest
density; two stage calcination the highest. This paper describes
the two-stage calcination process, which is most commonly used.
In this process, iron (II) sulfate heptahydrate is initially dehydrated
to a monohydrate: FeSO4•7H2O®FeSO4•H2O + 6H2O.
Subsequently the monohydrate is thermally decomposed at
temperatures exceeding 650°C to yield Fe 2O3. Final color depends
on the particle size formed and is controlled via regulation of the
time and temperature of the calcination process.
Iron oxide yellow, [a-FeO(OH)], is manufactured in colors ranging
from light yellows to dark bluffs. Final color is determined by particle
size that is generally between .1-.8mm. Due to their resistance to
alkalis, oxide yellow is often used to color cement. Three commercial
processes are used; the Penniman-Zoph process, the precipitation
process, and the Laux process.
The Penniman-Zoph process requires the preparation of ferrous
sulfate nucleating “seeds” or particles and subsequent oxidation
to ferric oxide. Initially ferrous sulfate is reacted at low temperatures
to yield the hydroxide: FeSO4 + 2NaOH®Fe(OH)2 + Na2SO4
The hydroxide is then oxidized to yield the hydrated ferric oxide:
2Fe(OH) 2 + .5 O 2®Fe 2O 3•H 2O + H 2O. Ferric Oxide seeds are
transferred to reaction tanks containing ferrous sulfate and scrap
4
iron in solution. The mixture is heated to between 70-90°C, air added,
and the solution circulated to initiate seed growth.
The following reactions occur during this process:
4FeSO4 + 6H2O + O2®4FeO(OH) + 4H2SO4
4H2SO4 + 8 FeSO4 + 2O2®4Fe2(SO4)3 + 4H2O
4Fe2(SO4)3 + 4Fe®12FeSO 4
The net reaction being 4Fe + 3O 2 + 2H2O®4FeO(OH). Process
variables which directly effect product quality and determine the
shade of yellow include the temperature of the reaction, the pH of
the solution, the circulation rate of oxygen, the circulation rate of
the ferrous sulfate solution, and the size and shape of the seed
particles. The longer the reaction is allowed to proceed, the larger
the particle sizes developed and the deeper the color. The reaction
is terminated once the desirable hue is achieved. The precipitate is
washed of any residual soluble salts, dried, ground, and packaged.
The precipitation process involves the hydrolysis of ferric solutions
with alkalines such as sodium or calcium hydroxide. Ferrous salts,
which are commonly waste by-products from other metallurgical
reactions, are often used. Ferrous chloride or sulfate is oxidized
and the precipitate washed, separated by sedimentation, and dried.
Yellow oxide produced by the Laux process is actually a by-product
of the manufacture of aniline. By reacting nitrobenzene with iron
and water in the presence of aluminum or ferrous chlorides, high
quality iron pigments are produced. Aniline is filtered off and the
pigments separated from unreacted iron.
The Impor
tance of pH Measur
ement
Importance
Measurement
In titanium dioxide and iron oxide production, pH is a crucial
measurement. Control of process pH is among many process
variables which ultimately determines end-product quality. Particle
size, distribution, opacity, color, product purity, and production yields
can all be affected by pH. Maintaining proper pH is essential to
optimize oxidation-reduction states, achieve desired particle size
and color, and to selectively precipitate desired end products.
Production yields, efficiencies, and product purity are also directly
affected. Insuring accurate pH measurement is essential to plant
operation.
pH measurement is used throughout the sulfate process for titanium
dioxide production. However it is considered a crucial control point
in the product finishing area for both the sulfate and chlorine
processes. Inorganic coatings are selectively precipitated onto the
TiO2 particle to increase product durability and to achieve desired
pigmentary properties. Precipitation is controlled by carefully varying
the pH and temperature during the finishing process. Ideally pH
must be controlled within several tenths in order to precipitate the
hydrous oxide onto the TiO 2 substrate with the desired thickness
and uniformity without precipitating titanium dioxide. The finished
product is subsequently filtered, washed, and dried.
pH measurement is also essential in the manufacturing of both red
and yellow oxides. Red oxide is primarily derived from the calcination
5
of iron (II) sulfate hydrate. A key control point is measurement of
the pH during the dissolution of iron in sulfuric acid. Accurate
measurement in yellow oxide production is also extremely important.
Similar to TiO 2 finishing, pH and temperature are varied to achieve
the desired end product pigmentary properties. Color and particle
size is directly attributable to the process pH, temperature,
recirculation flow, and reaction time.
Application Challenges
Accurate, reliable measurement of pH in pigment manufacturing
applications brings unique challenges and frustrations. The process
conditions themselves are very harsh and not conducive to in-line
pH measurement. Processes are generally quite acidic, often
running at values below 0 pH. Due to their batch nature,
temperatures are often cycled or run at elevated levels. Entrained
solids, fine particulate, and high salt content characterize the
suspensions. Each of these factors leads to high maintenance labor
and material replacement costs. Common problems associated with
these measurements include reference junction poisoning, abrasion
of the pH measuring glass, sensor drift, probe breakage, probe
fouling, constant cleaning and calibration, and ultimately frequent
probe replacement. In such applications it is not unusual to replace
sensors within days of initial installation.
It is not uncommon in pigment manufacturing facilities to clean and/
or re-calibrate pH electrodes multiple times throughout the day. This
is further complicated by the means in which the sensors are
installed. Often pH electrodes are installed in situ via extension pipe
into a tank. This makes it awkward to retrieve the sensor for routine
maintenance. Because sensor replacement is so common, many
companies have adopted plug-in sensor connections to facilitate
quick sensor replacement. Unfortunately temperature compensation
is generally not built into this connector, and the resulting
measurement is not temperature compensated. Attempts have been
made to install the pH measurement sensor in slipstreams or bypass
lines to facilitate isolation of the sensor from the process during
replacement. However, these lines often become plugged resulting
in static flow.
Yet another problem is the batch nature of the process. Often the
temperature and pH of the process is cycled to achieve optimal
product quality and physical properties. Temperatures can exceed
the upper limitations of the sensor. The pH can reach values which
are so acidic that destruction of the pH sensor is inevitable. Once
the batch is completed and the reaction tank emptied, the sensor
is often left dry. Thus, in each of these scenarios, the sensors must
be manually removed from the process and reinserted as process
conditions allow. This further adds to the associated costs of the
measurement point.
6
Following is an approximation of expected annual costs associated
with a single measurement point:
Typical Sensor Cost
Replacement Sensor Frequency
Yearly Sensor Cost
$ 225.00 each
2 weeks - 26 sensors/year
$5,850.00
Cleaning Frequency
Calibration Frequency
Associated Labor Costs
Once per shift
3 dai l y
15 minutes each -
Daily, 15 minutes each -
(assumes burden labor rate of
$50.00/hour, actual rates vary)
91.25 hours/year
273.75 hours/year
A n t ic ip at ed To t al Year ly Co s t
$18,250.00
$24,100.00
Application Solution
The ideal solution to the challenges presented with these
applications is to provide a sensor that has infinite life and requires
no maintenance. Unfortunately this can not be realized with existing
technology. However, a solution can be provided which drastically
increases the useful sensor life, while reducing the overall
maintenance requirements.
Automated pH holders are not new to the market. The concept is
quite simple. Provide a means of automatically controlling the
introduction and removal of a pH sensor into and out of the process.
Add to this the capability to automatically clean and assess the
sensor health both in situ and while removed. While simple in
concept, in actuality the same rigorous process conditions that affect
existing sensor measurement performance impact the holder design
criteria. In addition, a suitable pH electrode must be supplied which
provides accurate measurement under existing chemical, thermal,
and mechanical conditions.
Several basic system design conditions must be met. The holder
must permit the electrode to be withdrawn and reinserted into the
process automatically without interruption of the process. For safety
reasons, only holders equipped with limit switches for end position
monitoring should be used. If power supply, air supply, or
mechanical fault were to occur within the holder, it could remain
undetected and consequentially result in costly damage. For this
reason, the right electrode holder materials and seals for each
application must be carefully selected. Most problems encountered
with automatic pH measuring points are related to electrode holder
wear. Additionally, the holder must not execute any undesirable
movements in the event of a power or compressed air supply failure.
Following a power failure, the system must automatically initiate a
start-up in a predefined manner. When a pH electrode does fail, it
must be easily removable from the holder without allowing leakage
of process solution to the outside.
7
The primary task of the pH measuring analyzer is to analyze and
display the pH value based on the signal from the pH electrode.
However, a number of additional functions are being increasingly
integrated into the measuring analyzer. In addition to monitoring
the electrode glass membrane breakage and fouling, they must
include alarm and control functions. Given this growing complexity,
measuring instruments with a plain-text user interface should be
chosen to simplify operation and minimize the risk of operator error.
In order to increase accuracy, the analyzer should have both Nernst
temperature compensation and solution temperature compensation.
The latter should be used whenever calibration is performed, since
buffer solutions are temperature dependent. In most cases,
automated pH systems require an electro-pneumatic control cabinet
that executes the control commands from the analyzer. To assure
reliable operation of the entire system, all components of the system
must be integrated.
Retractable pH Holder
For the pigment applications
noted, a pneumatic holder
constructed of Kynar (PVDF),
PEEK, Titanium, or Hastelloy is
used. A microprocessor based
controller/transmitter
and
pneumatic interface provide
integrated system control and
measurement output to the
distributed control system. An
integral ball valve built into the
holder provides isolation of the
sensor from the process during
cleaning, calibration, and routine
sensor replacement. Cleaning
cycles are programmed into the controller for local control and binary
coded inputs are used to additionally allow the distributed control
system to remotely retract, insert, or initiate a sensor cleaning as
process conditions require. Pressure sensors incorporated into the
design of the holder provide feedback regarding holder position
(inserted/retracted). Interlocks prevent reintroduction of the holder
insertion shaft with the sensor
removed. A retraction security lock
insures that process pressure does
not force the sensor from the
process, should loss of system air
occur. An exter nal injector
integrated into the system allows
introduction of rinse water and
cleaning agents into the wash/rinse
chamber of the holder. A typical
cleaning cycle is described in the
diagram.
Additional cleaning programs
available include interval cleaning,
interval measurement, and a
weekly program. The rinse,
8
cleaning, and dwell period set points are adjusted as needed
depending on the batch operation requirements. Utilities required
include 120 VAC, rinse water (29-145 psi), and dry,oil-free
compressed air (60-90 psi).
The pH sensor used is a gel filled, 12 mm, pH sensor with 100 ohm
RTD and single Teflon reference junction rated to a working pressure
of 90 psig and maximum temperature of 130°C. An optional liquid
filled sensor with flowing zirconium reference junction and external
electrolyte reservoir has also been used due to the high salt
concentration and fine particulate materials found in the process.
Sensor diagnostics and holder problems are detected and indicated
on the analyzer display. When servicing is required, individual
components can be easily replaced because of the system’s
modular design. A representative system is shown below.
The holder itself is installed on a
reaction or finishing tank
Typical Automatic Cleaning pH Measurement System
recirculation line. Operating
temperatures and pH values
vary throughout the production
cycle depending upon the
desired end product. In yellow
iron
oxide
production,
temperatures range fro m
40-60°C and pH is maintained at
specific set points between 1-4
pH depending upon the time in
the reaction cycle. In TiO 2
finishing, the temperature is also
cycled and pH variations are
controlled between 5-7. Sensors
are manually calibrated prior to
the initiation of each batch. In the
yellow oxide [a-FeO(OH)] process, sensors are initially retracted
from the process as the pH is less than -2 pH. Rather than
continuously monitoring pH, the sensor is introduced at
predetermined intervals into the process as measurement and
control data is required. Sensors are also retracted as reactor
temperatures exceeding the rating of the sensor are reached and
upon completion of the batch. This technique increases average
sensor replacement intervals from a matter of days or weeks to
months. Automated cleaning of the sensor reduces manual
maintenance intervention and calibration requirements. This is
especially important in TiO2 finishing where solids content can
approach 60%. Multiple daily manual cleanings are instead replaced
by semiannual, scheduled recommended maintenance of the holder.
Calibration frequency is also decreased as sensor life and
performance is increased.
The net benefit to plant operations is realized in instrumentation
reliability, enhanced confidence in the accuracy of the measurement
point, and decreased operating and maintenance costs. Although
significant savings are achieved in material replacement costs, the
greatest savings come from reduced maintenance requirements.
9
Following is an example of expected initial equipment costs and
approximated annual costs associated with a single automated
measurement point:
Required Hardware
Approximate Costs
Retractable Kynar pH Electrode Holder
System Controller and Analyzer
Sensor
Injector
Cables
Installation
$ 5,530.00
5,345.00
194.00
1,215.00
180.00
400.00
A p p r o x im at e In s t alled Har d war e Co s t
Expected Annual Mainteneance
Replacement Sensor Frequency
Yearly Sensor Costs
Maintenance Parts
$ 12,864.00
Estimated Costs
4 per year
each sensor $ 194.00
(annual o-ring replacement) 1,100.00
A s s o c iat ed Year ly Co s t s
$ 1,876.00
Quarterly (8 hours/year for sensor
replacement and holder maintenance)
Weekly (15 minutes each)
13 hours/year
Maintenance Frequency
Maintenance Labor
Calibration Frequency
Calibration Labor
A s s o c iat ed Year ly Co s t s
(21 hours at $50.00/hour) $ 1,050.00
TOTA L Year ly Co s t s
$ 2,926.00
The calculated return on investment of the automated, retractable
system versus the fixed measuring point installation is as follows:
First year savings (Labor and Materials)
$ 21,174.00 ($24,100 - 2,926)
Initial System Hardware Costs
$ 12,864.00
Ret u r n On Inv es t m en t
~ 7.3 m o n t h s
The required hardware and maintenance part prices are actual
manufacturer list prices. Estimated labor requirements are indicative
of typical automated pH measuring systems in these applications.
Actual associated labor costs will vary depending on the actual
plant and are provided merely as an estimate. Regardless, the
estimated initial costs show a return on investment of approximately
7.3 months. Future savings are estimated at nearly $21,000.00 per
year per measurement point. When multiplied by the number of
potentially applicable measurement locations within a typical
pigment plant, annual savings can be substantial. Additional savings
and benefits to the plant are realized in improving product quality
through tighter process control and measurement accuracy.
10
In conclusion, automated, retractable pH systems provide a solution
in the process measurement of pH in difficult pigment applications.
Flexible system design allows the operator to tailor the operation to
increase sensor life, reduce maintenance requirements, and improve
measurement accuracy. Net benefits to the plant are realized in
reduced maintenance and operating costs as well as increased
product quality and tighter process control.
Actual holder installation in TIO2
finishing tank application
11