The role of sodium silicate in the flotation deinking of newsprint at Mondi Merebank

Jimmy Pauck and Professor Jeremy Marsh


The deinking process employed at Merebank is outlined and illustrated. A brief literature review of the relevant pulping and flotation chemistry is presented, with particular reference to the role played by sodium silicate. Sodium silicate's role appears to be multi-faceted and in some dispute. The mill's own experience has shown that the sodium silicate plays a vital role in the deinking process.

Sodium silicate's ability to disperse ink, both alone and in the presence of calcium ions and fatty acid soaps, was investigated using a model ink system. A representative newsprint ink base was dispersed under a variety of laboratory conditions and the resulting dispersions were studied. Sodium silicate proved to be a poor disperser of ink particles, but nevertheless appeared to greatly influence the dispersing properties of the soap in the presence of calcium ions.

The nature of the interactions between sodium silicate, calcium ions and the collector soap is being studied in an attempt to elucidate the role of sodium silicate. Although the work is not complete, early indications are that the sodium silicate competes with the soaps for calcium ions by virtue of the common ion effect. This results in a higher concentration of soap in the pulping process than would be the case if sodium silicate were not present. It is postulated that this would lead to improved dispersion of the ink and hence improved removal in the flotation process.


The Merebank De-inking Process
The recycled fibre plant at Mondi Paper's Merebank mill was built in 1990. The plant consists of equipment supplied mainly by LaMort and Voith.

The plant recycles 85 000 tons per annum of waste paper and produces about 70 000 tons per annum of deinked newsprint pulp. This constitutes about 25% of the newsprint furnish of the two newsprint machines at the Merebank mill.

The waste paper is collected and sorted off-site by Mondi Paper Waste and delivered to the mill, baled as "flat news" and "magazine".

A wide variety of printing inks can be encountered in a de-inking plant. In our own particular case, mainly offset printing inks both as cold-set and heat-set are involved.

A typical composition of these inks is presented in Table 1 below (1).

Table 1 – Composition of Off-Set Printing Inks

Table 1

Once the ink has been applied to the paper in the printing process, and drying has taken place, the pigments and resins remain behind whilst the mineral oil vehicle has either flashed off or penetrated into the fibres. The component that must be removed by the de-inking process consists of a matrix of pigment, which is mainly carbon black, and hydrocarbon resin binder. These components are by their nature very hydrophobic.

Figure 1

Figure 1 – Schematic Diagram Of The Deinking Plant

The layout of the plant is illustrated in Figure 1. The process is essentially a single stage alkaline flotation process, followed by a washing stage, disperser and hydrosulphite bleaching. A 70/30 blend of newsprint and magazine paper is fed into the pulper, together with 10.0 kg/ton (kg active chemical per ton of dry pulp) of sodium silicate, 4.8 kg/ton hydrogen peroxide, a trace of enzyme scavenger and enough caustic soda to bring the pH into the range 9.8 - 10.2. The waste is pulped at a consistency of 10-15% and at a temperature of 45 oC. for about 15 minutes.

After pulping, the stock is diluted to a consistency of approximately 7% before the poire, 4% before the HD cleaners and 3% before the coarse screens. After screening further dilution to about 1% takes place, and 4 kg/ton of collector soap and 2.0-3.5 kg/ton of calcium chloride are added just before the flotation cell. The target calcium hardness in the flotation cell is 240 ppm as calcium carbonate. The ink is floated off in a two-stage Voith flotation cell, at a pH of 8.2 - 8 .5. Thereafter the pulp is again cleaned, screened, thickened and washed, before it is bleached at medium consistency with sodium hydrosulphite, to a brightness of over 62. At the washing stage sulphuric acid is used to adjust the pH of the pulp to 6 -6.5. The water recovered in the disc filter thickener (decker) is split into two streams, designated cloudy and clear. The cloudy water is heated to 50 oC. with steam and used for dilution in the pulper and up to the flotation cell. From the flotation cell forward the clear water is used for dilution. Excess soap and calcium hardness will be circulated back into the pulper from the disc filter. After the disc filter, white water from the paper machine is used for washing and dilution of the pulp.

Process Chemicals and their Functions

The main chemicals added to the process are summarised in Table 2 below.

Table 2 – Overview of Process Chemicals and Components

Table 2

Sodium silicate's role appears to be multi-faceted. In addition to its hydrogen peroxide stabilising action, and its buffering and saponification properties, it has been reported to assist in the dispersion of the ink particles and influence their size (17,18,19). It appears to act as a collector (20), it reduces fibre losses and suppresses the flotation of fillers (7,21,22). However, a number of authors refute these findings and claim that it has no influence on the final brightness (12,22,23). The mill's own experience has shown that the sodium silicate plays a vital role in the deinking process. The elimination of sodium silicate from the process resulted in an immediate deterioration in deinking performance (24).

Sodium silicate, when dissolved in water at high pH, would consist in large part of the silicate anion (SiO4 4-)( 25) . The silicate anion would exist either on its own, or as a polymer, depending on concentration and pH. In either case it is a strongly ionic, highly polar molecule with no hydrophobic character. Although it seems unlikely that such a molecule could disperse a highly non-polar hydrophobic entity such as an ink particle, there are a number of references in the literature to the ability of sodium silicate to disperse the ink particles in the pulping process (2,3,17,18,19). In addition, sodium silicate's dispersing properties are well known in the field of detergents (25).


A considerable amount of work has been reported in the literature on the effects of sodium silicate on the de-inking process, some of which has shown contradictory results, as discussed in the section above. However, nothing has been reported in the literature as to the mechanism of these effects. This work is an attempt to begin to explain how sodium silicate works in a deinking system. As a first step, it was decided to investigate the dispersing behaviour of sodium silicate in more detail, in order to determine if it played a role in the deinking process. The deinking system as it occurs in a typical commercial process is very complex. In order to simplify the study it was decided to use a model ink system to study the interactions of sodium silicate with the other components in the system. These methods have been employed by other workers (26,27), but most of this work was done with flexographic inks, which are easily dispersible in water.

Dispersing Properties of Sodium Silicate
A local supplier of printing inks was approached to supply a sample of cold-set offset ink. The supplier agreed to supply the ink base only, excluding the mineral oil solvents. The solvents are dissipated in the printing process and the remaining pigment and binder is what must be removed from the fibres in the de-inking process. Working with an ink base would provide a material that closely resembles the composition of an actual ink particle. A method was developed that could produce a stable ink dispersion. Briefly, the method consisted of dispersing the ink in a hot aqueous solution under vigorous agitation, followed by ultrasonic dispersion. The amount of dispersed ink was determined by measuring the absorbance of the dispersion at 555 nanometers. The ink dispersion was filtered through a Millipore filter. The particle size distribution of the ink particles remaining on the filter was analysed by image analysis. The individual ink particles were isolated and examined under a scanning electron microscope (SEM). Electron dispersive spectroscopy (EDS) was performed on the individual particles to determine their surface composition.

A number of dispersing systems (incorporating sodium silicate, soap and calcium in various ratios) were compared at addition levels of 0,1,2,and 4 g/l of sodium silicate. This range encompasses the practical range of addition of sodium silicate in commercial systems. At 0 g/l of sodium silicate, caustic soda was used to adjust the pH of the solution into the range 10 to 11. Commercial grades of sodium silicate (79oTw.) and soap were used in all the work.

A number of practical difficulties were experienced and although every care was taken to reproduce the amount of dispersing energy, the repeatability of the results was not all that good. Nevertheless, the results of this exploratory work permitted some tentative conclusions to be drawn about the dispersing behaviour of sodium silicate. It was found that:

  • Sodium silicate does tend to disperse the ink slightly better than sodium hydroxide on its own. This dispersing ability increases with increasing silicate concentration and is influenced by the SiO2:Na2O ratio.
  • Sodium silicate does not disperse ink nearly as well as the soap.
  • The presence of the calcium ion, either with silicate alone or in combination with the soap decreases the dispersing ability.
  • Sodium silicate greatly modifies the dispersing behaviour of the soap, whether it is in the soluble sodium form or the insoluble calcium form.

The SEM micrographs showed particles that were spherical with more or less smooth surfaces. There was no evidence of colloidal silica on the surface of the particles, as has been proposed as a stabilising mechanism by some workers 20. The particles that were dispersed in silicate solutions alone showed some traces of silicon in the EDS spectra, but it is most likely that this was sodium silicate that had been occluded in the ink particle when it was dispersed. It appears as if sodium silicate itself has little dispersing action on the types of ink that occur in the Merebank deinking process. Nevertheless the dispersing tests clearly demonstrated that sodium silicate does influence the particle size and dispersing behaviour of the other main components on the pulping process, namely the calcium ion and the soap. The nature of these interactions needs to be investigated more fully. This investigation forms the second part of the experimental work reported in this paper.

Interactions between Sodium Silicate, Sodium Soap and The Calcium 2+ Ion
A series of two-component systems were used to investigate the interactions between calcium and sodium silicate and between calcium and the sodium collector soap. Calcium solutions containing 20, 50 and 150 ppm of Ca2+ were made up using the commercial grade of calcium chloride in use in the plant. The above range of calcium is representative of those concentrations which occur in the pulping process.

Sodium silicate and sodium soap solutions were added to the calcium solution in increasing concentrations and the pH was adjusted into the range 9,95 – 10,05 with sodium hydroxide or hydrochloric acid. The solutions were allowed to stand overnight to allow the precipitates that formed to settle out. The solutions were filtered if necessary and the clear supernatant solution or filtrate was analyzed for residual calcium. Calcium was determined by the classical EDTA titration method, using Calgon indicator to indicate the endpoint. Residual calcium concentrations were plotted against the ppm of sodium silicate or soap added .

The Influence of Concentration
The residual concentrations of calcium remaining in solution after addition of varying amounts of sodium silicate or sodium soap are shown Figure 2.1 to 2.3 below.

Ffigure 2.1

Figure 2.1 – Residual Ca2+ Concentration in solution at a 20 ppm Initial Calcium Concentration.

Figure 2.2

Figure 2.2 – Residual Ca2+ Concentration in Solution at a 50 ppm Initial Calcium Concentration.

Figure 2.3

Figure 2.3 – Residual Ca2+ Concentration in Solution at a 150 ppm Initial Calcium Concentration.

At low concentrations of calcium (Fig. 2.1), the soap removes more calcium from solution than sodium silicate. However, as the concentration of calcium increases, the difference in calcium removal decreases, and at 150 ppm (Fig. 2.3), the sodium silicate becomes marginally more effective than the soap in removing calcium from solution.

The Influence of pH
The pulping stage of the de-inking process normally takes place at about pH 10. The calcium-silicate and calcium–soap interactions were studied over a broad pH range. The initial calcium concentrations were set at 50 and 150 ppm, and a 1:1 molar ratio of calcium to silicate/soap was maintained. The pH was adjusted using hydrochloric acid or sodium hydroxide. Figures 2.4 and 2.5 depict the calcium removal by sodium silicate and sodium soap at 2 levels of initial calcium concentration.

Figure 2.4

Figure 2.4 – Calcium Removal by Sodium Silicate and Sodium Soap as a Function of pH, at an Initial Calcium Concentration of 150 ppm.

Figure 2.5

Figure 2.5 – Calcium Removal by Sodium Silicate and Sodium Soap at an Initial Calcium Concentration of 50 ppm.

At lower initial Ca2+ concentrations (Fig. 2.5), soap is more effective in removing the Ca2+. At higher initial calcium concentrations, the soap is more effective at lower pH's, but as the pH value increases the silicate becomes more effective in calcium removal (see Fig. 2.4).

In order to determine whether the removal of calcium from solution was in fact due to the addition of soap or silicate or merely an effect of pH, the calcium concentration of a 150 ppm Ca2+ solution was studied over a pH range. Figure 2.6 indicates that Ca 2+ would not be removed from solution under pH 11, which is above the pH normally encountered in the deinking process.

Figure 2.6

Figure 2.6 – The pH Behaviour of Calcium.

The Influence of Temperature
All of the experimental work was carried out at room temperature, about 23 oC. However, pulping is normally carried out at about 45 oC. The effect of temperature was determined at pH 10 and at 1:1 molar ratios of calcium to sodium silicate and soap. Figure 2.7 and 2.8 depict the calcium removal from solution by sodium silicate and sodium soap as a function of temperature, at initial calcium concentrations of 50 ppm and 150 ppm respectively.

Figure 2.7

Figure 2.7 – Calcium Removal by Sodium Silicate and Sodium Soap as a Function of Temperature, at an Initial Calcium Concentration of 50 ppm.

Figure 2.8

Figure 2.8 – Calcium Removal from Solution by Soap and Sodium Silicate as a function of Temperature, at 150 ppm Initial Calcium concentration.

Both figures 2.7 and 2.8 indicate that overall, temperature does not play a significant role in calcium removal.


Both sodium silicate and soap react with calcium to form insoluble precipitates. At low concentrations the soap tends to react preferentially with the calcium, but as the calcium concentration increases the sodium silicate exerts increasing competition for the calcium ion, until the reaction with silicate eventually predominates.

Whereas the ability of sodium silicate to remove calcium ions from solution increases with increasing pH, this is more pronounced at higher concentrations. In addition Ca(OH)2 only starts to precipitate out at pH values above 11. Therefore, the insolubility of calcium hydroxide is not a contributing factor to calcium removal under normal pulping conditions. In addition, the ability of soap to remove calcium ions from solution is not greatly influenced by pH.

The influence of temperature on the calcium ion removal by either soap or silicate is not great. All the experimental work was carried out at room temperature, whereas the pulping reaction occurs at elevated temperatures, typically at about 45 oC. The temperature profiles shown in Figures 2.7 and 2.8 indicate that any conclusions drawn from working at room temperature should apply equally well at the higher temperatures.


The studies of the dispersing behavior of sodium silicate using model ink systems has suggested that sodium silicate has very little affinity for the ink-water interface and on its own has only a small dispersing power for typical offset newsprint inks. However, its presence in a system of ink, calcium and soap has a big influence on the final dispersion result.

A study of the interactions of calcium with sodium silicate and the soap commonly used in a de-inking system showed that both the soap and the sodium silicate compete for the calcium ion, producing insoluble precipitates in the process . The soap has a greater tendency to react with calcium at low concentrations, whereas the sodium silicate tends to predominate at higher calcium concentrations. The effect of pH is most significant on calcium silicate, and the effect of moderate temperature elevation is also not great.

The pulping stage of the deinking process can be considered to be essentially a dispersing process, or in the words of Dobias et. al (1), a "come-off" process. The paper fed into the pulper is broken down under mechanical action into discreet fibres and filler particles. The ink particles are separated from the fibre surface by a variety of mechanical and chemical actions and dispersed in the water phase. If the ink particles are not adequately stabilised in the water phase, they can re-deposit onto the fibre surfaces with an adverse effect on the final deinking result. On the other hand, the flotation phase of the deinking process is essentially an agglomeration process, whereby the ink is agglomerated into larger particles and removed from the system by preferential adsorption at the water- air interface created by a bubble rising through the pulp mass in the flotation cell.

Calcium in the form of the Ca2+ ion is essential for this agglomeration process to occur, but it has been shown that calcium in solution is detrimental to the pulping process 1 .

High hardness in the pulper produces an inferior de-inking result. Calcium is therefore detrimental in the first (pulping) stage of the process, but essential in the second (flotation). It is also important to note that as the pulp passes from the pulper to the flotation cell, a huge change in chemistry occurs. An approximate ten times dilution takes place, along with a pH drop from 10 to 8.5. Under these conditions, the silicate "releases" the calcium and its sequestering effect decreases, thereby making the calcium available for its role as a collector in the flotation process.

It is proposed that one of the most important roles that sodium silicate plays in the deinking process is that it reacts with the calcium in the pulper, and in so doing effectively lowers the calcium hardness in situ. This allows the soap molecules present in the aqueous phase to more effectively disperse and stabilise the ink particles, because the soap is free from the interfering precipitation effect of the calcium ion. This phenomenon has been used in the detergent industry, and is known as "building" (25).

It is apparent from the literature that sodium silicate's role as a "calcium manager" in the de-inking process has not been fully appreciated. Once this role is more fully understood, it should be possible to more effectively manage the calcium in the deinking system, thereby improving the final deinking result and reducing the over-usage of chemical additives.


Further work on the three-component silicate/calcium/soap system is in progress, and early results are indicating that increasing levels of sodium silicate produce higher levels of soap in solution. Furthermore, it is proposed to transfer this work onto a laboratory scale pilot plant, to confirm that the proposed interactions do in fact occur in a real system, and to use this knowledge to optimise the chemical additions to the process.


I would like to thank my employer, Mondi Paper, for affording me the opportunity to study for an M.Sc.; Mr John Hunt for his help and encouragement; the laboratory staff at the Merebank mill for their assistance; to Coates Bros. (SA) Ltd . for supplying the samples of printing ink and to Dr. Fiona Graham of the Electron Microscopy Unit at the University of Natal for her assistance with the SEM and EDS studies.


1. DOBIAS, B., KLAR, W., SCHWINGER, K., Flotation of pigments and inks from waste paper. Innovations in Flotation Technology, 499-511, 1992.
2. FERGUSON, L. D., Deinking chemistry part 1. TAPPI J., 75(7):75(1992)
3. FERGUSON, L. D., The role of pulper chemistry in deinking., TAPPI 1991 Pulping Conference Proceedings, TAPPI Press, Atlanta.
4. McCORMICK, D., Chemistry of flotation and washing for deinking newsprint part ii. TAPPI 1991 Papermakers Conference Proceedings, TAPPI Press, Atlanta
5. BORCHARD, J., K., Chemistry of Unit Operations in Paper Deinking Mills. Plastics, Rubber and Paper Recycling 1995, 323-41, ACS Symposium Series.
6. BORCHARD, J., K., The use of surfactants in deinking paper for paper recycling. Current Opinion in Colloid & Interface Science 1997, 2(4), 402-408.
7. TURVEY, R., W., The role of calcium ions in flotation deinking. PITA Annual Conference 90 – Technology of Secondary Fibre, Mar 1990, Manchester, UK.
8. BAUMGARTEN, H., L., PTS Deinking Symposium, Feb 1995, Muenchen.
9. SCHWINGER, K., DOBIAS, B., The influence of calcium ions on the loss of fibre in the flotation deinking process. CPPA 1991 Recycling Forum Proceedings, pp 73-79.
10. TURVEY, R., W., Why do fibres float? CPPA 1991 1st Research Forum on Recycling, Oct. 1991, Toronto, Canada.
11. LIPHARD et al. The surface chemical aspects of filler flotation in waste paper recycling. CPPA1991 Recycling Forum Proceedings, pp55-64..
12. ZABALA, J., M., and MCCOOL, M., A., Deinking at Papelera Peninsular and the philosophy of deinking system design. TAPPI J., 71(8):62-8(1988)
13. SHRIVER, K., E., BINGHAM, S., J., FRAZIER, M., W., The function of clay in flotation deinking. TAPPI 1990 Pulpng Conference Proceedings, TAPPI Press, Atlanta.
14. GRANT, J., BLAIN, T., Short sequence recycling: The factors affecting ink agglomeration and stability. Proceedings of the 3rd Research forum on Recycling 1995. CPPA., Montreal, Que., Canada. pp 247-255.
15. SHEN, J., ABUBAKAR, S., SPRINGER A., M., The influence of clay addition on flotation deinking. TAPPI 1995 Recycling Symposium Proceedings, TAPPI Press, Atlanta.
16. LETCHER & SUTMAN. The effects of magazine and filler on the flotation deinking of ONP. CPPA 1991 Recycling Forum Proceedings, pp 65-72.
17. ALI, T., McLELLAN, F., ADIWINATA, J., MAY, M., et al. Functional and performance characteristics of soluble silicates in deinking, part 1. CPPA 1991 Recycling Forum Proceedings, pp 87-93.
18. MAHAGAONKAR, M., BANHAM, P., STACK, K., The effects of different furnishes and flotation conditions on the deinking of newsprint. Progress in Paper Recycling, Feb. 1997, pp 50-58.
19. READ, B., R., Deinking chemicals and their effects. North Western Branch Papers 1986 No. 2. Institution of Chemical Engineers.
20. SANTOS, A., CARRE, B., RORING, A., Contribution to a better understanding of the basic mechanisms involved in the pulping and flotation of offset ink and particles. TAPPI 1996 Recycling Symposium Proceedings, TAPPI Press, Atlanta.
21. LIPHARD, M., SCHRECK, B., HORNFECK, K., Interfacial studies and application tests of the flotation of printing inks and fillers. 77th Annual Meeting, Technical Section, CPPA, Jan./Feb. 1991.
22. MATHUR, I., Preferred method for the removal of filler from deinked pulp. TAPPI 1994 Recycling Symposium Proceedings, pp53-58, TAPPI Press, Atlanta.
23. MAK, N., STEVENS, J., Characteristics of fatty acid as an effective flotation deinking collector. Proceedings 2nd Research Forum on Recycling 1993, CPPA, Quebec.
24. CROSBY, J., Private discussions with Mr. J Crosby, superintendent of the Mondi Merebank deinking plant. 1998.
25. Crosfield Technical Information, Issue No. 1, July 1993. Soluble silicates and their applications.
26. MILANOVA, E., DORRIS, G., M., Flotation of model inks part 1: Experimental methods. 79th Annual Meeting, Technical Section, CPPA, A85-A95.
27. DORRIS, G., M., NGUYEN, N., Flotation of model inks part 2: Flexo. Ink dispersions without fibres. 2nd Research Forum on Recycling (1993).