Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
'-- 2 1 97256
PROCESS FOR IMPROVING THE DlSPERSlBlLI~ OF POST-TREATED
TITANlUM DIOXIDE INTO ORGANIC SYSTEMS INCLUDING PAINTS,
TEXTILES AND POLYETHYLENE - BASED MASTER BATCHES AND
ORGANIC SYSTEMS CONTAINING SUCH POST-I'REATED TITANIUM DIOXIDE
CROSS RF.FF.RF.~CF TO RFr ~TFn APP~ ~CATION
This application is a United States counterpart application to ori~in~ting German Patent
Application No. 19607249.2 filed February 27,1996, the disclosure of which is hereby incorporated
by reference.
FlF.T.n OF THF. INV~.~ITION
The invention is an improved process for enh~ncing the dispersion properties of post-treated
titanium dioxide pigment into a wide variety of organic and other systems incl~lding paints, textile
fibers and plastic such as polyethylene. While particularly directed to tit~ni--m dioxide pl~,lll~..l~i made
by the s~lph~te process, the invention is useful for pi~ments made by either of the two process types
of titanium ~ Yi~le, i.e. chloride as well as sulfate.
DESCRIPTION OF THE PRIOR ART
Po~ also rerelled to as after-lleAllll~lt or coating, oftit~nillm dioxide pi~~ to
improve the pigment's dispersibility into organic systems such as paints and plastics has long been
known. In the case of high-grade titanium dioxide pigments, the coating of the titanium dioxide
particles is generally followed in commercial practice by a classifying-milling operation whereby
agglomerates coming out the coating and drying process apparatus undergo particle size reduction.
See, for example, U . S . Patent No. ',846, 1 5 1 .
21 ~7256
. . . ~
U.S. Patent No. 2,032,827 discloses a mill in which a flow of gas is produced transversely to
the flight direction of particles. causing the residence time and the degree of comminution to be
changed. A jet mill of this kind also has the property, therefore, of classifying in accordance with
particle size, which can also be described as sifting.
European Patent No. 0 035 076 proposes that the coarse fraction in the aftertreated pigment
~ undergo size reduction by subj~lin8 the pigment, prior to the afte~LIç~ e~.~ step, to wet milling and
wet classification. In the course of this procedure it has been found that there is a considerable
worsening in the filterability of the aftertreated suspension.
SUMMARY OF THE INVENTION
Titanium dioxide has long been used as a pigment in organically-based paints, textile fibers
and plastics inc~ ng polyethylene and polyethylene master batches. Improving the dis~el ~;l)iliL~ of
the titanium dioxide pigment into such organic systems has long been a goal of customers who are
mon--f~ctu~er3 of D~uch ~Dh...3. Impro~ing titamum dioxidc by titanium dioxide manufact~ to
make their pigment more dis~e.~;l,le, to satisfy this customer desire, is an ongoing industry-wide
research goal in the titanium dioxide industry.
It has been found that. in a process of producing treated tit~ni~m dioxide pigments, their
dispersibility can be improved and the coarse fractions normally found in present day commercial
pigment can be reduced sufficiently, if the tit~ni~lm dioxide pigment is classified by a defined separate
air classifier apparatus in the dry state prior to the post-treatment step.
21 9/25~
DESCRIPTION OF THE PREFERRED EMBODIMENTS
The invention hereof relates to a process for making titanium dioxide pigment which
comprises carrying out, prior to post-treatment of such pigment, a separate step of dynamic
classifying of the tit~nillm dioxide with an air classifier.
Although at first glance it may be perceived as disadvantageous that an additional apparatus
is used for separate classifyine of the drv pigment prior to post-treatment, for the post-lrea~mel~l
procedure as a whole~ the invention hereof is found to be very favorable and to be economically
advantageous in terms of the improved pigment quality obtained.
In accordance with a variant of the novel process of this invention, untreated coarse titanium
dioxide particles (i.e. particles the majority of which are about 3 ~lm or greater, which are tli~tinctly
greater than the mean p;~5 n~lll particle size of 0.2 - 0.3 ~m) are removed and recycled and, together
with additional titanium dioxide starting material not yet milled, are supplied back to the mill for
further milling. As a result, the effective throughput in aftel ll~ll.lellt is increased and there is no
need for coating or post-lr~ 1 of the coarse particles which, following fur;ther c~ .. in
prior art processes, must be aftertreated and coated once again.
A st:~ar~le air classifier apparatus has been un~Ape~;ledly found to be far more efficient than
a classifying apparatus integrated in a mill to achieve sufficiently sharp classification. Such apparatus
make it relatively simple to alter cutoffpoints.
It has been surprisinL~ly discovered. furthermore, that the removal of oversized particles by
means of air classifvine prior to a~te~t~eat~l,enl is often sufficient by itself to produce readily
dispersible pi~ments makin~ it possible~ in some circumstances. to omit customary downstream steam
millin.~ (with inte~rated classifvinl~). It is even possible to obtain a pi~ment having ~ood
21 Y7256
dispersion properties a~er drying and without additional milling, for exarnple after a spray drying
step.
The separate classifying unit used in the instant invention is a dynamic air stream classifier.
The air stream classifier separates into two or more classes the mixture of"raw" tit~nillm dioxide
particles by their grain size and weight, with one of the classes being decign~ted "coarse". The air
stream classifier can be matched to the requirements and properties of the particular tit~nillm dioxide
to be treated. Air classifier devices are described in Ullman, Encylcopedia of Industrial Chemistry.
Vol. B2, pages 17.1 to 17.17.
Since titanium dioxide has a tendency to fomm deposits, the path of the product using this
invention avoids dead spaces, constrictions and deflections of material, (preferably by means of the
air classifier's flutter pipes, vibration sha~s, and movable guide vanes). The tit~nillm dioxide to be
classified is sufficiently p~ ~d.s~,cl ~ed, which is ensured by the classifying app~ al~ls use of air. Such
lng is followed by air classifying, where the coarse titanium dioxide particles are separated
offon the basis of their higher mass, by mearLs of c~ ;rJ8~l force in, in a ~,~f~ d embodiment, a
helical flow.
Particularly useful for this invention are air classifiers made by Alpine A.G., Augsburg,
Germany, particularly Model 400, and by Condux Maschinenbau Gmbh, Hanau, Germany,
particularly Condux's Superfine Classifier Model CFS.
Vaned classifier rotors are used in air cl~csifier designs. The ~low properties between the vanes
and in the vane-free intemal area of the rotor is important. The flow between the vanes is generally
steadv when entry is shock-free. The forrnation of turbulence between the vanes is detected on non-
shock-free entrv. The flow in the vane-free intemal area is correctly described as vorte~c flow - with
21 97256
the exception of the boundary layer on the front surface.
Therefore - under the conditions assumed - many classifiers have a classifier rotor in which
classification takes place between the vanes while in others the internal vane-free area is used for
separation.
The latter provides several practical advantages:
~ The conditions of flow entry can be selected relatively freely since the condition of
shock-free entry does not need to be fulfilled. The combinalion with a mill, for
instance, can be realized without any problems.
~ Same fineness can be obtained with a lower rotor peripheral speed when the unit is
designed accordingly. In particular, this can increase time of rotor wear. and
durability.
When de,;~ned for rl~csifiç~tion in the vane-free internal area, the flow in the gap between
the vanes ll ~ISpOl IS the feed material inwardly and outwardly (particles that are too large must be
tra~s~)olled out ofthe ~ P~t wheel again). The vanes have the task of accel~dt~ , the e.lt~...~
air stream loaded with material, if possible without losses, to the pe~ i?hc~ ~I speed of the rotor at the
vane's inner edge.
To ~ n..~ize the effects of delay in the boundary layers on the front surfaces of the classifier
rotor, the use of an immersion tube~ fitted in the discharge opening for the fine material and classifier
air~ has proved to be successful. Without wishing to be bound by theory, the effect of the immersion
tube can be explained by a vortex created on the side of the immersion tube facing the classifier zone.
This vortex prevents coarse oversized titanium dioxide particles carried inwards in the boundary layer
from entering the fine material discharge.
21 97256
Finally, it can be assumed, that even when the classifier rotor is designed for "internal
cl~c~ifi~tion", plelhllil~y rl~csific~tion will take place at the outer edge of the vane, thus preventing
titanium dioxide particles that are much too coarse being drawn in to the rotor. This assumption
explains why rotors of this design operate almost independent from the product load of the
classification air.
~ Typical appala~u~s pararneters employed for this invention are as follows:
Separation limit do 8 - 80 I,lm
Input Quantity 500 - 1500 kglh
Speed l O00 - 4000 rpm
EXAMPLES
The invention is described in more detail below by way of example.
F.Y~m,ple 1
This e~ elll used as a starting material titanium dioxide calcil-er discharge directly
o~tAined from the sulfiate proce~s. A R~ .ont Mill with a ~static clsc~fi~ wa~ used to IFint the
titanium dioxide p.;7~ 1 Aluminum or ~Illminum and silicon oxide hydlaLes were de?o~;led on the
titanium dioxide using known inorganic aftel ll ~,~l .u~ -t methods.
Two sets of samples were made, one using the technique of this invention. The novel titanium
dioxide pi~ differed from prior art pigments only in that they have been subjected to additional
classifying with a dvnamic air classifier (a MPW 400 made by Alpine A.G.) prior to aftell,e~ll"ent.
Final milling was then carried out for both sets of samples in a spiral jet mill (a steam mill) 200 mm
in diameter using high-pressure steam with the addition of trimethylolpropane (TMP) or trimethylol-
ethane (TME) as a grinding aid.
21 972~i
The dispersibility of the titanium dioxide pigments obtained was tested in accordance with
methods similar to those used in technical and scientific practice for the technology involved, i.e. in
systems in which the pigments described are typically employed. Table I lists in each case various
well known parameters which are a measure of the dispersibility of tit~ni~1m dioxide pigment.
In the case of the dispersion test in polyethylene (PE), a 50% titanium dioxide conce..l~te
(masterbatch) was prepared.
The sieve residue SR was used to assess the degree of dispersion of a titanium dioxide
pigment in the concentrate. The investigations with the 34 !lm sieve were carried out with a
Brabender measuring extruder under the following test conditions:
Barrel temperatures:
~one I 1 70~C
Zone 2 200~C
Zone 3 200~C
Screw speed 120 n~in-'
Plotter speed 10 cm/h
Sieve 34 ~lm mesh size
At the beginning of each experiment~ 400 g of PE-LD, Melt Flow Index (MFI) = '0, were
extruded throu~h the sieve at a screw speed of 120 min-' until a constant pressure Pl (horizontal line
on pressure plot) became established. Without interrupting the extrusion, I kg of the respective 50%
TiO. concentrate was extruded through the sieve at the same screw speed. When the entire amount
of concentrate has been e~truded throu~h the sieve. the extruder was run to empty for a short time
so that the pressure suddenlv fell and. consequentlv, a marked final pressure P. could be read off.
21 ~-/256
A measure of the pressure increase of I kg of MB (50% TiO,) is given by:
~P= P~ ~ P,
If this value is multiplied by 2, the result is the pressure difference for I kg of tit~nillm dioxide
pigment in bar/kg TiO..
A~er each experiment, extruder and sieve assembly were flushed through with unpigmented
polyethylene and cleaned. A~er di~ aJIlLn~ of the sieve, a new sieve assembly was employed for the
subsequent experiment and operation continued as described initially.
A second separate ~lycol S !lm test was prefonT~ed as described hereafter. The ~Iycol 5 ~,lm
sieving test is a method of determining the filterability of a 20% TiO. suspension in ethylene glycol.
The lower the passage time or the higher the ptl~ ~nlage of filtrate in this test, the better is the
dispersibility of the pigment. A mixture of 60 parts by weight of tit~ni~lm dioxide pigment and 40
parts by weight of ethylene glycol were milled for 30 minutes in a bead milt (Netzsch, PE 075). The
f~t~ g suspen~ion is dihtect with further ~ e glycol untit the content of titanium dioxide i~ 20%
by weight. 500 ml of this su~pena;on were passed under suction, with the aid of a waterjet pump, at
col,~ underpressure through a metal sieve (5 ~,lm mesh size, 5.5 cm in diameter). The propollion
ofthe su~,ue-,~;on which passes throu~ and the time required for it to do so, was determined for this
filtration.
21 ~7256
Exemplary data obtained are compiled in Table 1.
Table I
Base substance* unclassified ¦ classified nn~ ssified ¦ classified
A~el~eal~"e,ll 0.6% AI,03 0.9% SiO, 1.5% AI,0
Pressure increase amount sieved ~g] 333 1000 154 718
34 ~,lm sieve bar/kg TiO. 1617 193 3682 689
Sieving residue mg/kg TiO~ 40 21Sieve fracture 28
40!1m sieve
S ~m sievingtest filtrationtime [s] n.d. 37 53 29
filterable fraction [%3n.d. 100 65 100
*Classified is the inventive example.
Discussion of Results:
When reviewing the two pigment suspensions/masterbatches, which differ only in the
pigments used (classified or unclassified), the throughput of the suspension/masterbatch with the
ccified F;~ 1 is in a ro~ ~e~b~e way higher, the p,~ i"~,~ cj~ifi~ntly lower and the
sieved - residue less than in the suspension/masterbatch with ~Incl~ssified pigment - all these
measurements show in the suspension/masterbatch with classified pigment a smaller quantity of
unwanted oversized particles.
The same conclusion is apparent from the S ~,lm sieving test results. If a suspension contains
less large particles (=classified)~ filtration is quicker. filtration timer is lower and the filterable quantity
higher
-10-
21 9725~
Example 2
The same rnilled base substance as in Example 1, without previous inorganic afte. l,ea1ment,
was subjected to an organic aftertreatment in a steam mill as in Example 1, the organic substances
used being trimethylolethane (T~E), neopentylglycol (NPG) or trimethylolpropane (TMP). Table
2 again compares the corresponding measurements of the prior art pigment and of novel material
~ which, prior to a~ertreatment, has been classified in the same apparatus as in Exarnple 1.
Table 2
Basc Substancc* llnrl~csifi~d ¦ cl~ccified Imrl~c5irlt-d ¦ cl~csificd llnfl:lccifi~d ¦ cl~csifi~d
Ancrlrca~rncn~ TME NPG TMP
Prcssurc ~ncrcasc arnounl sic-cd fgl l90 1000 200 1000 1~7 1000
34 ~lm sicvc bar/kg TiO. 2542 224 2538 203 2620 198
Sie- ing residue mg/kg TiO. 92 18 97 22 70 20
40 ~,Im sieve
5 ~lm sieving ~ filtration time Isl 52 26 53 27 62 28
filterable fraction 1%1 100 100 100 100 81 100
*Classified is the inventive example.
Discussion of Results:
In E~cample I the aRell-edtl-lent was ino~al1ic (standard). It is shown by the results in Table
. that the claimed et~ect is observable in the same way on pigments using well-~;nown organic
a~ertreatment substances. The same distinct effects (with the same explanation) as in E~ample 2 are
observed.
