Pigment Dispersion with Surface-Active Agents - Industrial

May 1, 2002 - Earl K. Fischer, and Charles W. Jerome. Ind. Eng. Chem. , 1943, 35 (3), pp 336–342. DOI: 10.1021/ie50399a017. Publication Date: March ...
0 downloads 0 Views 1MB Size
JIEN’T DISPEl ION with

-Aetive A EARL K. FISCHER CHARLES W. JEROME Interchemical Corporation, New York, N. Y.

The theoretical basis for the use of surface-active agents has been extensively discussed in the literature. The assumption usually made is that a monomolecular film, adsorbed and oriented a t the interface of the solid and the liquid, produces a marked transition in the properties at the interface (4, 9,16,18, go, 29).

F THE large number of reagents now marketed as surface-active agents, many are of distinct utility in the dispersion of pigments in the nonaqueous media used as vehicles in the paint and printing ink industries. Industrial uses for these compounds in aqueous systems are extraordinarily varied (I, 7 , 10, 16, 23, 27). That extensive application is possible for surface-active agents in nonaqueous systems is becoming apparent from a growing number of patents relating to specific products and processes as well as papers describing experiments ( 2 , 3 ) . The addition of small quantities of reagents to provide desirable working properties for paints and printing inks is not an innovation, for the practice extends back to the origins of the industry. Commonly used materials include soaps, tallow, stearin pitches, maxes, gilsonite, etc. (28). While the function of these substances is neither entirely clear nor simple, one important result of their addition is the change in flow properties of the composition, rendering it either more fluid or contributing to an increase in consistency or body. The synthetic compounds now commercially available greatly extend the number of reagents available for incorporation in pigment dispersions. The object of this work was to make a systematic survey of the surface-active agents on the market with reference to the manner in which they affected the dispersion of pigments and to obtain data from which fairly general deductions could be drawn. Representative surface-active agents were selected from each of the several classes of commercial species available. The effect of these compounds on typical pigment vehicle combinations was studied in relation to the pigment and vehicle types, concentration, and method of addition. The rate of dispersion and final tinting strength of the mixtures were measured since these factors are important, However, it was the change in rheological properties of the dispersion that was taken as the primary physical criterion of the effect of a given surface-active agent. The results obtained have been an aid in selecting surface-active agents for specific formulations of analogous systems.

0

Definitions The term “surface-active agent” has come into use as a general designation for compounds which in small quantities modify the properties of a system by adsorption at an interface. The terms “wetting agent” and “dispersing agent” are frequently used synonomously, and although some writers attempt a distinction there is no general agreement on special meanings. The terms used in this paper follow the general usage in paint and printing ink technology as noted below (8, il, 13, Wi, 22). Pigment powders are considered as aggregates or masses of homogeneous particles which require an appreciable amount of mechanical work to separate the component particles. “Aggregates” are to be distinguished from “floccule$” and the state of “flocculation” where the combination of pigment particles is disrupted by weak mechanical forces or by a change in the interfacial chemical forces, “Deflocculation” represents the state in which the pigment particles are independent or unclustered. The “ultimate working unit” may be an individual pigment particle or group of particles SO firmly held together that they remain intact during a commercial application. “Grinding” is the process by which pignent aggregates are reduced to the size of ultimate pigment particles, with an implied reference to some device for the application of mechanical work. “Dispersion”, a generic term, refers to any process, procedure, or state relating to heterogeneous systems of solids and immiscible liquids. I n the paint and printing ink industries a pigment-vehicle system in which the pigment is finely divided and deflocculated is considered without exception a good dispersion. 336

INDUSTRIAL AND ENGINEERING CHEMISTRY

March, 1943

Where aggregates are present in the disperse phase, even though the particles are deflocculated, the material is usually considered a poor dispersion. There is less agreement in nomenclature, however, where the disperse phase is finely divided but highly flocculated. It is in this last case that surface-active agents are of greatest possible utility.

Flocculation and Yield Value The flocculation and wettability of pigments in lowviscosity suspensions were studied in detail by Ryan, Harkins, and Gans (18). Solids which are deflocculated settle slowly to small final volumes; flocculation is shown by rapid settling to high final volumes. Where the disperse phase is present in higher concentration, the flocculated structure leads to plastic flow and measurements of Eectly useful for the evaluation of surface-active

t

Deflocculated FIGURE 1. DISPBRSION OF TITANIUM DIOXIDEIN LmSEIED V A R N I S H ( X 1450)

Flocculated --t

The effect of surface-active agents on pigment dispersions has been studied to aid in formulating general principles for commercially important products such as printing inks and paints. Because of the number of combinations possible, the experiments were limited to major types of pigment-vehicle systems as determined by basic wetting characteristics. Well characterized surface-active agents were selected, each representative of a class of compounds. Additional variables included dispersion procedures, concentrations, pigment-vehicle ratios, and methods of addition. Data are given on the

337

agents. The connection between flocculation and plastic flow, as measured by plastic viscosity and yield value, has been shown by Green and others (6,19,19,2b), The photomicrographs of Figure 1illustrate a deflocculated and flocculated dispersion of titanium dioxide in linseed varnish. The yield value of the deflocculated dispersion is low, 300 dynes per sq. cm.; that of the flocculated dispersion, identical i n formula except for the addition of a trace of flocculating agent, is of the order of 35,000 dynes per sq. om. The former is fluid, the latter is a semisolid. I n the evaluation of surface-active agents for pigment dispersion, the change in flow properties of the dispersed system thus may be taken as the primary criterion for the effectiveness of the reagent, measurable as the magnitude of the change of plastic viscosity and yield value. The control of flow properties or consistency is essential both in the manufacturing and use of many products. For many purposes it is desirable to decrease the yield value of a dispersion, resulting in a fluid composition from a combination initially highly plastic. Deflocculation also permits an increase in the quantity of the disperse phase with retention of useful flow properties. On the other hand, it is sometimes desirable to increase the yield value of a pigment dispersion, and this result may sometimes be obtained by the proper selection of surface-active agents.

Plan of Experiment Because of the countless combinations of pigments, vehicles, and surface-active agents, a complete survey was neither feasible nor necessary. Instead, experiments were limited to representative types of dispersions, selected on the basis of wetting characteristics. For this purpose pigments considered as hydrophobic, neutral, and hydrophilic were chosen. Vehicles included glycerol, a Litho varnish, mineral oil, and a varnish composed of a phenolic resin in mineral oil solvent. Surface-active agents were selected to include a t least one of each of the major classes; preference was shown for compounds of fairly well characterized chemical identity and uniformity rather than for proprietary compounds of unknown or undisclosed composition. Data in this paper are taken from a large compilation of experimental results obtained during the last six years. I n addition, information from experiments on other aspects of the subject of pigment dispersion-for example, the

rate of strength development as a measure of dispersion rate and the rheological properties of the compositions. While a degree of specific action is associated with each reagent, several criteria have been established to aid in the selection of surface-active agents for special purposes. The degree of flocculation, indicated by measurements of plastic viscosity and yield value, can be materially affected by surface-active agents. The rate of pigment strength development appears to be a secondary effect related to the change in plastic viscosity. The utility of surfaceactive agents in this field is indicated.

INDUSTRIAL A N D E N G I N E E R I N G C H E M I S T R Y

338

TABLE I. SURFACE-ACTIVE AGENTSUSED Suppliera Trade Name Type Aerosol O T

Aresket 300

Dioctyl ester of sodium sulfosuccinic acid Monobutyl diphenyl sodium monosulfo-

nate

Benzidine

Organic amine

Copper oleate

Soap Polymerized sodium salt of alkyl aryl sulfonic acid Sodium salt of lauryl Duponol M E sulfate Gilsonite (Selects) Bitumen Daxad 23

Ink Lengthener EE Lecithin Xacconol NR Santomerse D

Sapamine KW

Tergitol7 Zinc naphthenate

Petroleum residue Plant phosphatide Sodium alkyl aryl sulfonate Alkyl aryl sulfonate Diethyl aminoethyl stearylamide hydroacetate Sodium salt of sec-alcohol sulfate Soap

American Cyanamid & Chemical Corp.

Monsanto Chemical co. National Aniline & Chemical Co., Inc, Shepherd Chemical Co. Dewey and Almy Chemical Co. E. I. du Pont de Nemows & Co., Inc. Allied Asphalt & Mineral Corp.

Sun Oil Co. W. A. Cleary Corp.

National Aniline

& Chemical Co., Inc. Monsanto Chemical

c o.

Ciba Co., Inc. Carbide & Carbon Chemicals Corp. Nuodex Products Co., Inc.

a Compounds similar to those listed are available from other suppliers under different trade names. The substances used were considered as examples and are but a small part of the totalnumber tested. Inclusion of a given substance in the list should not be construed as a specific recommendation for that product.

direct transfer of aqueous pigment pulps to oil dispersion (flushed colors)-have contributed to an understanding of the role of surface-active agents, but space Liniitations prevent the inclusion of all relevant data. The surface-active agents tested are identified as completely as possible in Table I Some of the compounds are suitable only for dark-colored products; for carbon black a number have become well known-e. g., benzidine and copper oleate (24, 26). The pigments mere typical commercial products. Toluidine red is an azo dyestuff prepared by coupling diazotized m-nitro-ptoluidine with P-naphthol. Barium lithol toner is the barium salt of the compound formed by coupling diazotized @-naphthylamine-&-sulfonic acid (Tobias acid) with @-naphthol, Representative of the larger class of organic colors, toluidine toner and barium lithol toner may be considered hydrophobic pigments. Iron blue, of the type known commonly as Milori blue, is prepared by the oxidation of ferrous ferrocyanide, and for simplicity may be considered as ferric ferrocyanide. Iron blue is considered neutral in wetting properties (6), but some commercial species are definitely hydrophilic. Ultramarine, considered hydrophilic, is the synthetic pigment obtained by the reaction of sulfur, carbon, sodium carbonatc, and clay. The titanium dioxide used was a typical commercial material of anatase crystalline type. Carbon black or channel black, manufactured

Vol. 35, No. 3

by the combustion of natural gas, is considered hydrophobic or organophilic ( 6 ) , but' this pigment exhibits variable wetting characteristics depending on previous treatment and adsorbed moisture. Because the moisture content of pigments is variable and depends on storage conditions, all pigments were heated for 24 hours in a n oven a t the following temperatures immediately before use: Toluidine toner, barium lithol toner Iron blue Ultramarine blue, titanium dioxide, carbon black

65' C.

::1

The purpose of this drying schedule was to reduce variations in pigment properties without modifying the pigment character from its normal condition; pigments, especially carbon black, are often stored in a hot room before use. The vehicles used were as follows: Linseed or Litho varnish, obtained by the heat polymerization of linseed oil, was a dark-colored type with an acid number of 13, bodied to a viscosity of 5.9 poises, a t 30" C. Glycerol was the c. P. product. The mineral oil was a refined ink oil with a viscosity of 5.8 poises a t 30" C. The resin varnish contained 40 per cent modified phenolic resin in a mineral solvent; the viscosity was 14 poises a t 30" C. The dispersions were prepared, except where othervise noted, by hand mixing the ingredients, and grinding (five tight passes) on a three-roll laboratory mill. A sequence of operations was established so that all dispersions were prepared in the same way, including time elapsed between milling and examination, mill setting, and mill temperatures. Plastic viscosity measurements were made on rotational viscometers. In the first experiments two commercial viscometers were used. One was a Stormer equipped with a cylindrical bob and cup; the other was a hlacMichae1 viscometer equipped with a continuously variable speed transmission and lever counter. While both of these instruments were satisfactory for low-viscosity oils and pigment dispersions, they failed seriously on dispersions exhibiting highly thixotropic and plastic flow. The values obtained for printing inks were completely discordant, since plastic-viscosity and yield value are a function of the shearing stress and the attainment of constant thixotropic breakdown levels ( l a , 14). The Stormer, operated by weights, necessarily

FIGURE 2. GENERAL VIEW

OF THE VISCOlvlETER

INDUSTRSAL A N D E N G I N E E R I N G C H E M I S T R Y

March, 1943

imposes a variable maximum shearing stress, since the greatest weight, rather than the highest rotational speed of the bob is constant. The MacMichael viscometer was capable of higher shearing stresses at speeds to 240 r. p. m., but the maximum speed of the cup in this instrument was not even approximately reproducible for highly plastic dispersions. Variable shear, together with inadequate temperature control, made the values so obtained of relative significance only. The data of this paper were obtained on the viscometer described by Green ( I $ ) as constructed in these laboratories, which overcame the operational difficulties enumerated above, This instrument (Figure 2) embodies the essentials of precise machining and alignment, rugged construction, temperature control of * O . l o C., and accurately reproducible speed control. With this viscometer a measurement includes the following steps: The pi ment dispersion is placed in a cup, and after temperature equigbriurn is reached, the cup is started rotating at a ow speed. Readings are taken at increasing speeds to a maximum of 200 r. p. m. The cup is operated at this speed until equilibrium in thixotropic breakdown is attained (2 to 10 minutes) as indicated by an unchanging deflection on the torsion dial. Readings are then taken at decreasing speeds for eight to fifteen speed settings. A curve is obtained from the data by lotting revolutions per minute as the ordinate and degrees ofdeflection as the abscissa. The slope of the straight line part of the curve is a function of the plastic viscosity, and the intercept is a function of the yield value (Figure 3). The instrumental data were converted to poises and dynes per sq. om. by the following equations ( l a , 17):

U f

= 9.55 (T = T2C

339

21

II

t

12

a: LT 9 '

41

- Tz)S/r. p. m.

where U = plastic viscosity, poises f = yield value dynes per sq. cm. S = (l/R,"- 1/R2)/4rh C = S/ln(R,/Ra) 5" = torque Tt = intercept on torque axis Rb = radius of bob, cm. R, = radius of cup, cm. h = heightofcup,cm.

I

I

I

0 I4A0L DEFLECTION 80 -DEGREES I20

16(

FIGURE 3. VISCOMETER CURVES FOR ULTRAMARINE BLUEI N MINERAL OIL A , control, no reagent. B , plus Aerosol OT: C,plus Aresket 300. Solid circles f e resent readings a t increasing shearin stresses (up curve) ; open circ?es, a t decreasing shearing stresses (8own curve). Torque conversion: 1" = 96.2 dyne-om.)

All measurements \yere made at 30" * 0.1' C. The instrumental constants were determined both by direct calibration with torque measurements and calibration with viscosity reference standard oils obtained from the National Bureau of Standards. For the determination of tinting strengths, white bleaching inks were prepared by grinding zinc oxide in each of the vehicles used in the tests. The pigment dispersions were reduced with the appropriate white ink, and the deviation from standard strength, taken as 100 per cent, was observed visually and calculated fiom the relative proportions of white and tinting material. Concentration of Surface-Active Agents

A quantity of reagent estimated to form a monomolecular film on the pigment particle may be considered the minimum to register a change in properties of a dispersion. Calcula-

tions of this quantity for commercial pigments on the basis of a number of arbitrary assumptions were reported (2); while these figures are only estimates, they serve as a rough guide. For many pigments 1 per cent by weight is sufficient; for pigments of small particle size, such as organic toners and carbon black, 2-3 per cent is required. In one series of experiments the effect of concentration of different surface-active agents was determined. An observable change in yield value was obtained in most cases a t 0.5 to 1.0 per cent, with the maximum result a t approximately 3 per cent. For commercial applications the minimum useful quantity should be known; for the purposes of these tests, however, 3 per cent was used for all pigmentvehicle systems except carbon black where 5 per cent was employed. Large excesses of surface-active agents are of

-

TABLE11. EFFECTOF SURFACE-ACTIVE AGENTSON FLOW PROPERTIES OF ULTRAMARINE BLUEAND IRON BLUE

Reagent None (control) Aerosol O T Aresket 300 Daxad 23 Duponol ME Lecithin Nacconol N R Santomerse D Sapamine I