NC State
Hubbe, M. A. (2024). “Size press practices and formulations affecting paper properties and process efficiency: A Review,” BioResources 19(1), 1925-2002.


Size presses on paper machines are used to apply a solution of a polymer – usually starch – to the surface of the sheet and thereby to increase the stiffness, surface strength, and printing quality of the product. This article reviews publications dealing with the size press equipment, the materials, and factors affecting both the operating efficiency and attributes of the resulting paper. The emergence of film-press equipment (e.g. blade-metering size presses) in the 1980s has greatly decreased the frequency of web breaks and increased productivity. Starch technology at the size press, though relatively mature, continues to evolve. By adjustment of starch attributes, solids levels, and incorporating other additives, modern papermakers can tune size press outcomes to meet a range of paper product requirements, including strength, hydrophobicity, and the reduction of air permeability. By application of various synthetic polymers, mineral particles, and even nanocellulose in combination with starch or other base polymers, there is potential to extend the technology to meet a range of future needs for paper products.

Download PDF

Full Article

Size Press Practices and Formulations Affecting Paper Properties and Process Efficiency: A Review

Martin A. Hubbe

Size presses on paper machines are used to apply a solution of a polymer – usually starch – to the surface of the sheet and thereby to increase the stiffness, surface strength, and printing quality of the product. This article reviews publications dealing with the size press equipment, the materials, and factors affecting both the operating efficiency and attributes of the resulting paper. The emergence of film-press equipment (e.g. blade-metering size presses) in the 1980s has greatly decreased the frequency of web breaks and increased productivity. Starch technology at the size press, though relatively mature, continues to evolve. By adjustment of starch attributes, solids levels, and incorporating other additives, modern papermakers can tune size press outcomes to meet a range of paper product requirements, including strength, hydrophobicity, and the reduction of air permeability. By application of various synthetic polymers, mineral particles, and even nanocellulose in combination with starch or other base polymers, there is potential to extend the technology to meet a range of future needs for paper products.

DOI: 10.15376/biores.19.1.Hubbe

Keywords: Film press; Blade-metering; Starch; Viscosity; Hold-out; Paper stiffness; Paper strength

Contact information: Department of Forest Biomaterials, College of Natural Resources, North Carolina State University, Campus Box 8005, Raleigh, NC, 20695-8005, USA; Email:


This article reviews published information about size press practices, including the equipment, procedures, materials, operating efficiency, and effects on paper properties. The main goals of running a size press on a paper machine include increasing paper’s stiffness and surface strength. The resulting paper is usually less dusty, such that it can be printed with less contamination of printing presses. Such attributes mostly can be attributed to three main categories of factors, which will be considered in this review, namely the size press equipment, the applied materials, which traditionally have been forms of starch, and various procedural options.

The scope of this article can be defined, first of all, by describing how a size press differs from a conventional aqueous coating operation used in the manufacture of paper. Both operations apply aqueous formulations continuously to the surface or surfaces of a web of paper. A key difference is that the solids content (as the sum of dissolved and any insoluble components) is much lower in the case of size presses, which typically are run in a range of 5 to 18% of soluble polymeric matter (Klass 1998). By contrast, typical water-based coating formulations often have a total solids content in the range of 30 to 65% (Klass 1998). The major component of a typical coating formulation is insoluble pigment, e.g. clay or calcium carbonate. Although some papermakers apply mineral pigment at a size press, the major polymeric component usually still is dissolved starch. Another distinguishing feature is that a size press operation takes place during passage of the sheet between a pair of cylindrical rolls, whereas most coating operations involve a blade or some other device to remove an excess of applied material.

Unlike typical coatings on paper, a typical size press application is not expected to result in a continuous film (Cushing 1979). Due to the relatively high amount of water present in the formulation, there would not be sufficient material to completely fill the void spaces in the paper. Pan et al. (1995) developed a way to predict the progressive emptying of pores and deposition of material during the drying of paper coatings, and this mechanism seems to be appropriate for the situation being considered here.

Fig. 1. Depiction of expected main locations of the size press formulation before and after drying, such that the solids end up spread one the fiber surfaces and collected at junctions between fibers

Figure 1, which is based on the cited work, illustrates a hypothetical process by which evaporative drying after a typical size press can be expected to promote accumulation of starch and other size press components at or near to the junctions between fibers within the paper sheet, as well as on the outer surfaces of the fibers. In this figure the fibers are depicted in cross-sectional view, such that their central lumen spaces can be seen. Readers seeking information about aqueous coating technologies for paper are recommended to study appropriate parts of the textbook by Paltakari (2009).

Though many of the articles to be cited in this document show clear improvements in paper properties resulting from size press applications, it is important to keep in mind that any rewetting of a paper sheet can negatively affect the inherent properties of the base sheet. This was shown most clearly by Lepoutre et al. (1986), who applied water to the paper surface as one of their control conditions. The strength and other properties of those water-sized sheets were always inferior to those of the basestock. The effect was attributed to the swelling and debonding of the paper web. In addition, there is no wet-pressing operation after a size press.

Another way to define the scope of this article is to consider where the size press fits into the overall process of manufacturing a hypothetical paper product. The size press is almost always an “on-machine” unit operation, meaning that it operates in a continuous mode together with the paper forming device and associating paper drying equipment. Thus, the paper sheet entering the size press has just been formed, pressed, and then dried for the first time by passage around a series of steam-heated dryer cans (Smook 2016). Since the paper becomes wetted by the starch (or other) solution applied at the size press, it subsequently needs to be dried again, almost always by means of a further series of dryer cans. Typically, about 1/3 of the drying capacity on a paper machine comes after the size press (Eklund 1989). Depending on paper property requirements, the surface-sized paper could then be coated or subjected to various converting operations.

A key potential advantage provided by a size-press operation is the fact that nearly 100% of the added material goes directly onto or into the paper product (Helle 1971; Hoyland et al. 1977; Gray and Rende 2005; Hemmes and Wahl 2017). This is in contrast to the wet end of a paper machine, where certain components that are small or lacking in strong binding to the fiber surfaces may require multiple passes through the forming device before they are retained, and some may pass into the wastewater. An inherent disadvantage of size press usage is the fact that the paper needs to be dried twice. Adding a size-press operation to a paper machine will always increase the steam requirement. The increased drying requirement is likely to impose a further restriction on the speed of production. Another inherent issue concerning size-press application is that the distribution of the solution on the paper surface can be non-uniform. The uniformity of the applied size-press formulation also can be affected by base-sheet uniformity. The splitting of the starch film as the sheet exits the nip tends to produce an “orange peel” effect, the subtle nature of which often allows that effect to escape notice. However, when a dye or whitening agent is being applied onto to the surface, such effects can become apparent under appropriate lighting.

Fluorescent whitening agents (FWAs) provide a good example to illustrate the differing priorities when deciding to add something at the wet end of a paper machine, the size press, or both. At the wet end, where the priority is on achieving high retention onto the fiber surface, papermakers mainly employ tetrasulfonated FWAs (which are often the cheapest) or disulfonated FWAs (which are the least soluble and therefore the easiest to retain) (Sampl et al. 2021). By contrast, papermakers often employ hexasulfonated FWAs at the size press; they are the most water-soluble, which allows the material to diffuse and move throughout the sheet to some extent after application. The goal is to achieve a more uniform optical effect, especially when viewing under ultraviolet light.

After the main sections covering literature related to size press equipment and size press additives, some other emerging trends and opportunities will be reviewed, based on more recent publications. For instance, so-called “pigmented size press” operations have become much more attractive following developments in the design of size press equipment (which is the next topic). It has been proposed by some writers that by adding certain process chemicals at the size press, the wet-end chemical operations can be simplified (Brouwer 1997; Hemmers and Wahl 2017). In addition, the concept of adding nanocellulose at the size press offers some possibilities and challenges. Nanocellulose tends to increase the viscosity of a mixture, and excessively high viscosity can place a limit on the speed of running a size press. Finally, there is increasing interest in applying eco-friendly barrier materials at the surface of paper, so there will be strong interest in how size-press applications might contribute to such product goals.

It is important to acknowledge the existence of earlier review articles, without which the present article would have been much more difficult to complete. Some important sources are listed in Table 1.

Table 1. Review Articles, Books, and Chapters about Size Presses, the Applied Materials, and their Results


Three classes of equipment that can play critical roles in a size press operation are the size press itself, the run tank that supplies the formulation, and optional drying equipment, such as any non-contact drying devices. In addition, there are some important options to consider with respect to the devices used to prepare starch solutions.

Size Press Configurations

The most important distinction between different categories of size press devices has been between the traditional “flooded nip” size presses and the so-called “film presses” (Eklund 1989; Felder 1991; Kohl et al. 1999). In addition, certain types of coater devices sometimes have been employed in lieu of size presses to apply starch to paper, especially before the emergence of film presses (Fineman and Hoc 1978a,b; Klass 1988; Bailey 1996).

Typical size press equipment, including both flooded nip and film press varieties, involve passage of the paper web between two press rolls. In flooded nip (or “pond”) size presses, typically at least one of the rolls is “soft,” meaning that it is covered with a rubber or polyurethane layer (Moore 1998), and the other can be “hard,” i.e. a steel roll (Eklund 1989). Gray and Rende (2005) indicated that hardness values in the range of 20 to 30 P & J (measured with a Pusey & Jones Plastometer) are common. In principle, the compliance of the roll cover facilitates a more even pressure across the width of a paper machine. Coyle (1988) calculated the degree to which a soft roll surface can be expected to deform in the course of applying a fluid material to a non-porous web. The corresponding situation in which the fluid is able to penetrate into the web was considered by Ninness et al. (2000). Devisetti and Bousfield (2010) measured the pressure distributions when using porous paper webs with applications of fluids with press nips. Differential equations were developed to describe the deformation of the soft roll covers and the effects of such variables as fluid viscosity and speed.

Flooded nip size presses

In a flooded nip or “pond” type of size press, the paper is exposed to a pool of aqueous solution before entering the nip between two rolls. The press nip itself contributes to the role of metering the applied amount. As shown in Fig. 2, three main types of flooded nip size presses can be identified by the relative positions of the two press rolls, i.e. vertical, inclined, and horizontal (Eklund 1989; Smook 2016). From an idealized perspective, the horizontal format would seem ideal, since it would facilitate keeping equal conditions of wetting of the paper in the ponds on each side of the sheet. But the inclined format is most popular, probably because it is more consistent with the main direction of travel of the sheet from the wet end towards the dry end of the paper machine, and possibility minimizing the required space.

Fig. 2. Schematic depictions of three basic layouts of flooded nip size press that were in common usage before the general transition to film press systems

Before the introduction of film presses, one of the best options for starch application onto a paper web may have been the Billblade® coater, which is sketched in Fig. 3 (Hansson and Klass 1984; Klass 1988). Work by Fineman and Hoc (1978a,b) showed that such a coater could be used to upgrade the quality of printing papers. For instance, it was possible to reduce linting problems in offset printing presses. The linting problems were further reduced by use of a hydrophobic agent in the size press formulation.

Fig. 3. Schematic diagram of a Billblade® system, which was formerly sometimes used for starch application to paper

An inherent problem with flooded nip size press installations has been a relatively high frequency of sheet breaks. Such problems may become more serious with increasing web speed, which has been attributed to flow instabilities and splashing in the ponds (Linnonmaa and Trefz 2009). Exposing a paper sheet to a pool of relatively hot liquid incurs risk that the paper web will become wetted all the way to its core, thereby reducing its strength so much that the web breaks. Though such effects can be countered by hydrophobic sizing treatments during formation of the base-sheet, such internal sizing can raise the cost of production, and the web may still break due to defects. If the hydrophobicity or other properties of the paper are not equal across the width of the paper machine, the size press solution uptake may be non-uniform, which can lead to paper quality problems.

Although flooded nip size presses are still employed on some older and smaller paper machines, such practices are becoming less common. However, the distinction between flooded nip size presses and more recent generations of equipment needs to be kept in mind when reading literature about size press operations in general. As will be shown in subsequent sections, some operating variables that tend to be critical for flooded nip presses may be inconsequential for some other size press operations.

Pre-metered film presses

A film press can be defined as a type of size press in which the starch (or other) solution is first applied to one or both of the press rolls as a film, such that there is essentially no “pond” as the web passes through the nip. The rolls no longer have the function of metering the amount applied to the paper. Rather, the uptake is determined by (a) the amount of film applied to the roll, (b) the permeation into the sheet mainly due to the applied pressure in the nip (Eklund 1989), and (c) the splitting of the film that remains in mobilized form between the paper and the adjacent roll at the exit from the nip (Dobbs 1993). Because film presses, in general, do not wet the paper before it is actually passing through the nip, the possibility of complete wetting of the sheet is essentially eliminated, and this translates in a greatly decreased probability of size-press breaks of the web (Hiorns and Sharma 1996; Lipponen et al. 2004). A further contribution to this goal is provided by the fact that film press systems often allow for increases in starch solids (Lipponen et al. 2004). A higher solids level of the applied starch means less water to evaporate, in addition to a higher viscosity, such that the starch layer remains more towards the surface and the first layers of fibers in the sheet (Bergh and Hemmes 1991). Table 2 highlights articles focusing on a variety of aspects related to film-press operations and outcomes.

Table 2. Film Press Equipment and its Usage

Typical wetting times of the incoming sheet before the size press nip can be roughly estimated based on the process speed and the observed length of contact with any pond or contact with a film before the press nip. Note that a slower speed should be assumed in the case of a flooded nip size press, since it is unlikely that such equipment would currently be used on faster paper machines. Based on these rough estimates, there may be a ten-fold difference in pre-wetting time, if not more (Table 3).

Table 3. Estimates of Time of Wetting before the Nip, Depending on Wetted Length before the Nip and Operating Speed

Metering blade film presses

The general features of a metering blade size press are shown in simple form in Fig. 4, Part A. As shown in Part B, the applicator system is likely to be a type of “short dwell” coating system (Triantafillopoulos and Aidun 1990; Gray and Rende 2005), using equipment that is essentially the same as is commonly used for some pigmented coating operations directly onto paper. Short-dwell devices act like a nozzle, such that there is no contact between air and the starch solution before the spreading of the film. Regardless of how the film is spread, the application is made onto the surface of a relatively large rubber-covered roll (Eklund 1989). This type of device is well suited for the application of a starch film onto paper intended for printing purposes. According to Rantanen and Finch (1994), blade-metering presses also can be used for pigmented coatings, up to a medium level of solids.

Fig. 4. Schematic diagram of a blade-metering size press system. A: Basic layout; B: Option with a short-dwell applicator. The drawing of the short-dwell system was inspired by a similar drawing in Gray and Rende (2005).

Metering rod film presses

Metering rod systems are similar to what has been just described, except that the amount of size press formulation spread as a film is governed by what can pass through the openings between a wire-wound rod at the roll surface (Poranen and Kataja 2000). To change the amount applied, one can change either the solids concentration or replace the rod with one that has a different size of wound wire. Though the film applied to the roll may retain a fine streaky appearance, corresponding to the wrapped wires, the idea is that such features will become irrelevant as the film passes through the nip and gets partly squeezed into the paper surface. With respect mineral coatings, Rantanen and Finch (1994) recommend using groove roll applicators for film presses only for low-solids applications.

An alternative to the grooved rod is the smooth rod (Roper et al. 2022). The cited authors used such a device for the application of barrier coatings onto paper. When using a smooth rod, the amount of formulation applied to the press roll can be decreased by increasing the pressure, decreasing the diameter of the rod, or by decreasing the solids content of the formulation (Linnonmaa and Trefz 2009).

Gate-roll size press application

As a historical precursor to the modern film press, it is also possible to prepare a starch film or coating film on a transfer roll by means of a gate-roll application. This scenario is sketched in Fig. 5. As in the size press configuration depicted in Fig. 4, the gate role can be classified as a pre-metered system. Wilson (2005) reported that Beloit developed such a system in which only the two central press rolls rotated to match the velocity as the paper web. The next outer rolls rotated at 50 to 90% of the sheet velocity, and the outermost rolls rotated at 20 to 50% of the velocity at the outsides of the rolls. According to Klass (1998), gate-roll coaters were introduced in the 1960s, but they tended to suffer from short roll cover lifetime, problems due to film splitting, and limitations with respect to coating rheology. Linnonmaa and Trefz (2009) give 800 m/min as an upper limit for the speed during effective gate-roll usage for film press applications.

Fig. 5. Schematic diagram of a gate roll size press system

Jet applicators for the film press

As a more recent modification, jet applicators have been developed to gain better control and cleanliness during as a size press formulation is transported to the point of application onto the size press rolls (Kaipf et al. 2000). Rather than using a blade or rod, the formulation is applied by means of non-contacting multi-jet nozzles to the size press roll surfaces. Temperatures and flow rates are controlled. One of the reported benefits of this technology is decreased debris transferred to the surface of the first drying cylinder following the film press.

Coaters for starch application

Pigmented size-press formulations often have been used to precoat the paper sheet ahead of a final coating operation (Paltakari and Lehtinen 2009). Balzereit et al. (1995) ran trials and showed that a metering size press could be used to apply an initial pigmented coating (i.e. pre-coating) before a final blade coating. According to Bailey (1996) such applications can be expected to be used as pre-coatings, not for replacement of existing blade coating operations for existing grades of paper.


Run Tank for the Size Press

Before a starch solution is sent to the size press, it is usual that it to be held in a run tank (Walter 1998). This is where conditions such as temperature, formulation viscosity, pH, and other quantities can be monitored and sometimes controlled. Mixing within the run tank can ensure complete uniformity of the incoming starch solution with other additives, such as hydrophobic copolymers, colorants, fluorescent whiteners, defoamers, etc. (Paltakari and Lehtinen 2009). If there is any return flow from the size press (as is common for flooded nip presses), it will be a continuous input to the run tank too. Features of the run tank system can include level control, heating (sometimes by a steam jacket), solids or viscosity monitoring, and the filtering of the formulation. To minimize problems, a first priority is to avoid splashing and entrainment of air in the starch circulation system, especially in the path back to the run tank (Wilson 2005). Walter (1998) discusses how careful design of the run tank and related equipment can minimize the entrainment of air and the production of foam.

Dryer Adaptations for the Size Press

When applying starch solutions to the paper surface, it is typical to feed the outgoing damp web into a conventional dryer section, composed of steam-heated steel cylinders (Smook 2016). As was noted earlier, the proportion of “after-dryer” cans may be about 1/3 of the total on the whole paper machine (Eklund 1989), though that proportion might be subject to a decrease with an increase in solids level of the size press formulations. In principle, if there is a lot of wet solution at the paper surface, then there is a greater likelihood of starch deposition onto dryer cans; thus, conditions that favor more penetration of size-press formulation into the paper tend to reduce deposits on dryer can surfaces (Wilson 2005).

Other drying devices may be needed, at least initially, in cases where the size press formulation is tacky, as when latex products are in the formulation. In such cases it may be necessary to employ air-turns, infrared drying, or combinations of different non-contact drying technologies, at least initially (Rennes 1998a,b). For instance, Turunen (1996) recommended that non-contact drying be considered when running pigmented coatings at a size press. In addition, the multi-jet applicator system for film presses has been reported to result in less contamination of conventional after-drying operations.

Fig. 6. Possible positioning of online sensors to permit automatic control of drying operations associated with a size press (redrawn based on an original by Shapiro, 1998)

When preparing a range of paper products, with the use of a size press, there can be increased importance of maintaining tight control of operations. Shapiro (1998) showed that such control can be facilitated by use of certain online sensors. These can include sensors for temperature and percent solids. In addition, the pick-up amount can be controlled. A further step is to use a camera to detect possible defects after the operations. The likely positions of some of these sensors, related to drying of the web after a size press, are illustrated in Fig. 6.


Overview of Size Press Additives

Because starch plays such a large role in most size press operations (Cushing 1979), this section will first consider its main chemical properties, its most important sources, from the standpoint of papermaking, how it is separated from other components of the plant source, and how it is prepared for use at the size press. Other options include polyvinyl alcohol (PVOH) (Kane 1978; Lertsutthiwong et al. 2004; Kim et al. 2017; Bhardwaj and Bhardwaj 2018; Abhari et al. 2018a,b; Liu et al. 2021) and carboxymethyl cellulose (CMC) (Paltakari and Lehtinen 2009). Optional additives include hydrophobic agents, colorants, and fluorescent whiteners.

Starch Fundamental Properties


Regardless of the plant source, starch can be described as a polysaccharide composed exclusively of anhydroglucose subunits (Ogunsona et al. 2018; Cheng et al. 2021). Unlike cellulose, the anhydroglucose units are connected by the alpha rather than the beta form of glycosidic bonds, which leads to very different behavior. As shown in Fig. 7, there are two co-existing forms of starch that are present in most, but not all starchy plant materials (Bergh and Hemmes 1991).

Fig. 7. Two forms of starch macromolecules that coexist in most starchy plant materials.

A: Amylose, showing the linear chain structure and the native V helical conformation;

B: Amylopectin, showing the branched chain structure and linkage details

The simpler of the two is amylose, which is comprised of linear chains in which alpha glycosidic linkages connect the C1 of one unit to the C4 of the next. Depending on the plant source, the amylose often has a degree of polymerization of about 800 to 3000 (ca. 130,00 to 500,000 g/mole) (Ellis et al. 1998). The more complex form is amylopectin, in which approximately 4% of the anhydroglucose units provide a branch point, and these are associated mainly with its C6 hydroxyl group (Thompson 2000). The linear segments are typically about 10 to 12 anhydroglucose units in length. The high level of branching within amylopectin gives rise to relatively dense, compact molecular having a degree of polymerization of about 20,000 (i.e. molecular weights of about 300 million g/mole) (Willett et al. 1997).


An amylose molecule, when it has been dissolved in water, adopts various helical conformations (Imberty et al. 1991; Gessler et al. 1999; Tan et al. 2007; Putseys et al. 2010). The amylose chains may be present either as double helices (A- and B-type amylose) or as a helix formed from a single chain (V-amylose). The amylose components of native cereal starch, such as corn starch, tend to be dominated by A-amylose, whereas native tuber starch, including potato starch, tends to have more B-amylose (Imberty et al. 1991). The cited authors described how A-amylose can be converted to B-amylose.

The single-chain V-type helices have a central cavity that is relatively hydrophobic (Immel and Lichtenthaler 2000). The orientation of the –OH groups of the coils are such that they mainly render the outsides of the coils hydrophilic, whereas polar groups are absent from the insides of the helices. As a consequence, any hydrophobic components of the material, such as fatty acid monomers, are likely to be contained as inclusion compounds within the V coils (Yan et al. 2012). The cited study showed that palmitic and stearic acid could be included in size press formulations by this means, thus contributing to the hydrophobicity of a resulting linerboard product. It appears that the proportion of V-amylose increases in proportion to the amount of oleophilic monomers present (Cheetham and Tao 1998). X-ray crystallography has shown evidence of increasing V-amylose content in the crystalline phase with increasing proportions of amylose (Tan et al. 2007). However, in the cited work, double-helix forms of amylose tended to be dominant.


Within the plant, the amylose molecules are likely to be associated in a relatively loose crystalline arrangement of adjacent V-type coils (Immel and Lichtenthaler 2000; Conde-Petit et al. 2006) or double-helix coils. Amylopectin molecules, despite their complex structures, also have been reported to adopt a regular arrangement when present in plant materials (O’Sullivan and Perez 1999; Genkina et al. 2007). Though present together, the amylose and amylopectin may be present in different micro-domains within a starch grain.

Changes in crystal form

The state of crystallinity changes when starch is cooked, which is sometimes called a pasting operation (Wang et al. 2015b; Wani et al. 2016; Reyniers et al. 2020). The amylose form of starch generally shows the most significant effects of retrogradation (Miles et al. 1985; Fredriksson et al. 1998). Initially, the cooking appears to solubilize some of the amylose as intact V helices, as already described (Immel and Lichtenthaler 2000; Conde-Petit et al. 2006). In the case of amylopectin, the solubilized starch solutions appear to be amorphous and not readily subject to further formation of regular structures. Solubilized amylose from corn gradually adopts a double-helix form (A-helices) that can self-associate into undesirable structures, and these can separate from the solution (Conde-Petit et al. 2006; Fang et al. 2020). This process is called retrogradation or “set-back” by papermakers, and it generally hurts the ability of the starch to act as a bonding agent within paper (Liu et al. 2007). In general, retrogradation can be limited by applying the starch solution to paper soon after the starch has been prepared. In particular, long storage or gradual cooling of the starch solution is to be avoided (Cushing 1979). Walter suggests adjusting the temperature of cooked starch to no higher than 72 °C but no lower than 60 °C to minimize retrogradation. Retrogradation is also promoted by low pH and the presence of multivalent metal cations such as calcium or aluminum (Cushing 1979). Figure 8 illustrates one likely route of transformation among different helical forms of starch and their semi-crystalline agglomerated forms. Though amylopectin is generally not noted for a high level of retrogradation, Fechner et al. (2005) reported its conformational changes occurring relatively quickly after a pregelatinization process, by means of Raman spectroscopy. Blennow et al. (2001) showed that amylopectin can undergo agglomeration, leading to multi-chain structures, as quantified by gel permeation chromatography. Miles et al. (1985) described such changes to cooked amylopectin as being “reversible”.

Fig. 8. Simplified portrayal of changes in amylose starch conformation and colloidal form starting with the natural form present in the grain, then cooking, conversion to a different helical form, and finally retrogradation, which appears to involve aggregation of the A type of amylose helices

Starch Source Materials and Starch Isolation

Especially from a North American papermaking perspective, the most important plant sources of starch products are maize (corn) varieties, though potato starch is also widely used. Tapioca and wheat starches are also used, depending on the location (Ellis et al. 1998). Typical starch grain shapes and sizes are illustrated in Fig. 9 (Alvarez-Ramirez et al. 2019). In the cited work, the corn starch grains had a mean longest dimension of 12.2 µm, while the corresponding value for potato was 41.2 µm. In each case, it was determined that the grains swelled by only about 7% when immersed in room-temperature water for 8 hours.

Fig. 9. Basic starch grain sizes and shapes illustrated for two starch types commonly used in papermaking applications (traced from micrographs of Alvarez-Ramírez et al. 2019)

Corn starch and its isolation

The maize (corn) variety most often used as a source of starch for papermaking is Zea maize L., which has been called dent corn. This is the same species that is used for corn sweetener and other food uses; therefore, it has a huge volume of worldwide production. The mean diameter of grains is about 15 µm (Ellis et al. 1998), and the particles can be described as spheroidal, but not smooth. The term “polyhedral” has also been used (El Halal et al. 2019; Reyniers et al. 2020). The lipid content in a grain of corn starch is relatively high, about 0.6%, as is the level of protein (0.35%) (Ellis et al. 1998). The amylopectin level is about 72% of the starch, with amylose accounts for the remaining starch. The amylose in corn has an average degree of polymerization of about 800 (Ellis et al. 1998) or about 960 to 990 (Reyniers et al. 2020). Because its molecular mass is relatively low, compared to some other forms of starch, the amylose present in dent corn is more susceptible to retrogradation. This is consistent with the opaque nature of pastes and films prepared from dent corn starch, as would be expected from the inclusion of crystalline particles in the micrometer size range. The gelatinization temperature of dent corn starch is about 75 to 80 °C (Ellis et al. 1998).

Waxy maize is a hybrid form of starch that is grown exclusively for non-food applications, such as papermaking. Because it contains near to zero amylose, it is almost free from the effects of retrogradation after it is cooked. The lipid content (about 0.15%) is lower than that of dent corn. The starch grains are generally similar to those of dent corn, though the pasting temperature can be somewhat lower (65 to 70 °C rather than 75 to 80 °C) (Ellis et al. 1998).

To isolate maize starch, the grain is first soaked (steeped) in dilute sulfurous acid for one to two days (Yu and Moon 2022). The corn oil is extracted, and then the starch is separated from the gluten and fiber components following a wet-milling operation. Since wet-milling tends to release sulfur from the protein component, 1-cyseine can be added in an acidic medium as a reducing agent to release the starch from its protein binder in the grain (El Halal et al. 2019). These authors also described an alternative separation process starting with the steeping of corn kernels in 0.1% sodium bisulfite solution for about a day at 50 °C, followed by drainage and wet-crushing, filtration (100-mesh, then 270-mesh), decantation and removal of the supernatant, and then separation of the starch solids by centrifugation.

Potato starch and its isolation

Potato starch can be obtained from purpose-grown potatoes or as a byproduct from potato processing into such products as potato chips or French fries. Of the starch products used for papermaking, potato starch has the largest grain size (averaging about 30 µm diameter) and having a “fat oval” shape (Ellis et al. 1998; Reyniers et al. 2020). Potato starch typically has lower levels of lipids and proteins compared to maize, but it contains phosphorous (about 0.08%, dry mass basis), which gives it a weak anionic charge. The typical degree of polymerization of the amylose component in potato starch is about 3000, which is much higher than the other forms of starch (Ellis et al. 1998). This difference is expected to render the material somewhat slower to retrograde in comparison to maize and some other starches. Also, the pasting temperature (60 to 65 °C) is somewhat lower than the other forms of starch used by papermakers (Ellis et al. 1998).

Tapioca starch

Tapioca has similarities to maize starch, including its typical grain size, but its lipid and protein contents are lower, each at about 0.1% (Ellis et al. 1998). The degree of polymerization of its amylose component is relatively high, similar to that of potato starch, and its amylose content is relatively low, about 17% (Ellis et al. 1998).

Wheat starch

Wheat starch, in comparison to the types described above, tends to have relatively small grains (average diameter 10 µm) and relatively high levels of substances that might be regarded as impurities, i.e. 0.8% of lipids, 0.4% of proteins, and 0.06% of phosphorus (Ellis et al. 1998). The amylose content is as high as that of corn starch (28%), and the degree of polymerization of the amylose is also relatively low (about 800) (Ellis et al. 1998).

Starch Modification

For reasons of efficiency and economy, chemical modifications of starch are often carried out before the starch is either cooked (gelatinized) or converted (reduced in molecular mass). For that reason, it makes sense to describe the chemical modification steps first, even though they are optional, from the perspective of size-press practices. Chemical modifications can substantially raise the cost of starch production, and since the amount of starch applied at the size press can be relatively high (e.g., 5% of the mass of a typical printing paper sheet), there is a strong incentive to use underivatized starch, which is often called “pearl starch.” Some of the important chemically modified forms of starch are oxidized, cationic, and hydroxyethyl.

Oxidized starch

Oxidized starch products are widely used at the size press (Craig et al. 1968; Cushing 1979; Dobbs 1993). By reaction with sodium hypochlorite, some of the hydroxyl groups are converted to carboxylic acid groups, which contribute a negative charge to the starch when it its cooked. More detail related to oxidized starch is given later in the context of viscosity reduction, i.e. starch conversion.

Hydroxyethylated starch

One of the most widely used starch derivatives for size-press usage is the non-ionic product hydroxyethylated starch (Banks et al. 1973). This starch derivative is known for providing tough films. Though the derivatization necessarily adds to the cost of production, the resulting non-regularity of the product’s molecular structure renders it highly resistant to retrogradation (Ellis et al. 1998). The hydroxyethylated starch also has been reported to stay out near to the surface of paper (Cushing 1979). Such behavior is likely due to a reported tendency for transient self-association among the chains (Jauregui et al. 1995). Figure 10 shows the basic structure of hydroxyethylated starch. The relative frequency of substitution onto -OH groups in different positions within the starch molecular structure was studied by Merkus et al. (1977).

Fig. 10. Hydroxyethylated starch molecular structure

Cationic starch

Derivatization of starch with amine groups provides a positive ionic charge to the macromolecules when the starch is later dissolved. Though this can be considered almost an essential requirement for starch being added at the wet end of a paper machine (Harvey et al. 1979; Roberts et al. 1987), there are reasons for some papermakers to prefer cationic starch products also for size press addition. For one thing, if some defective paper needs to be repulped and formed again into paper (i.e. it is repulped), the cationic size press starch component will tend to be efficiently retained on the negatively charged surfaces of cellulosic fibers (Rankin et al. 1975; Cushing 1979; Hamerstrand et al. 1979; Ellis et al. 1998; Lee et al. 2002). In addition, it has been reported that cationic size press starch products have a greater tendency to stay out near to the surface of paper, thus providing more surface strength and stiffness to the product (Cushing 1979; Lee et al. 2002; Andersson and Järnström 2006). Though such effects might be attributed to electrostatic attraction of the cationic starch to negatively charged carboxylate groups at the fiber surfaces, it is also likely that differences in rheological properties are involved. Wilson (2005) mentions that early attempts in the 1950s to add cationic starch at the size press were unsuccessful, and it had been suggested that the problems were due to the electrostatic interactions with the fiber surfaces.


To prepare cationic starch, the suspension of starch grains is treated under highly alkaline conditions with epoxypropyl-trimethylammonium chloride or a related compound (Bergh and Hemmes 1991; Butrim et al. 2011). Subsequent cooking of such starch yields molecular chains that have an affinity for cellulosic surfaces.


An innovative technology has been demonstrated by which natural oils, or other hydrophobic monomeric compounds, can be incorporated into a specialty paper product, possibly allowing their gradual release over time (Aguado et al. 2022). This was achieved by crosslinking of the starch with cyclodextrin, taking advantage of the hydrophobic character of the interior of that ring-like saccharide. Note that this effect is mechanistically related to the hydrophobic nature of the interiors of V-type amylose helices, as described earlier (Immel and Lichtenthaler 2000).


For some specialty applications of size press starch, there may be an advantage of using a starch product that does not require cooking at the point of usage. Such advantages can be achieved by a process called pregelatinization (Alexander 1995; Liu et al. 2017). Briefly stated, a slurry of starch grains is heated to an optimized extent, then dried before storage and shipping. Alternatively, extrusion may be used as a means of achieving the same effects (Liu et al. 2017). Though pregelatinization does not immediately solubilize the grains, it renders them susceptible to subsequent swelling and solubilization when the redried grains are placed in warm water.

Starch Conversion

Solutions of native starch, at the concentrations that papermakers would prefer to employ them, generally have too high a viscosity to be able to pass correctly through a size press nip or through under the blades or rods used for applying the starch film to a film-press roll. Excessively high starch viscosity can cause “nip rejection,” which may manifest itself as unstable operation of the size press (Abell and Knowles 2001). In some cases, jets of starch solution are emitted from the pond area, possibly spattering the paper web (Wilson 2005). Thus, one of the first steps, carried out either at the starch company or in the paper mill, is to reduce the degree of polymerization to an optimum range suitable for the local equipment and the grade of paper being made. Conversion can be accomplished by use of enzymes, oxidizing agents, or ultrasonic treatments, as will be described below (Brenner et al. 2016).

Enzymatic conversion

When enzyme conversion of size-press starch came into use (Hughes and Craig 1950), it was viewed as an advantage, since it avoided the use of oxidizing agents (see later) (Bajpai 2018). According to Wang et al. (2022), enzyme-converted size press starch often achieves superior results compared to other forms of converted starches. The conversion is carried out with α-amylase, which is usually obtained from Bacillus subtilis. However, Wang et al. (2022) found advantages of using some different types of α-amylase to prepare size-press starch, possibly giving a better contribution to paper strength at a given level of viscosity. Conditions such as temperature and treatment time before usage need to be optimized, depending on the targets for viscosity and solids content (Li et al. 2013). The authors just cited suggest treatment at 65 °C for 20 minutes at 20% solids, with 0.02% enzyme dosage, a CaCl2 concentration of 0.04%, and pH near to 6. After the conversion, the starch needs to be cooked (see later) (Cushing 1979; Li et al. 2013; Wang et al. 2022). The overall procedure is shown schematically in Fig. 11. As an alternative, Cave and Adams (1968) describe a fully continuous process that can include both enzyme conversion and the cooking of starch.

Fig. 11. Steps in enzyme conversion of starch to decrease its degree of polymerization and viscosity at a given solid level. The term θ here refers to temperature.

Ammonium persulfate conversion

In-mill preparation of size-press starch is often conducted by treatment of the grain slurry with ammonium persulfate (Craig et al. 1968; Dobbs 1993). In addition to decreasing the degree of polymerization, the oxidation also induces an anionic charge to the starch, in its wet forms.

Hypochlorite oxidization conversion

Especially when the molecular mass is to be reduced before size-press starch is shipped to the paper mill, a preferred method is oxidation, by means of hypochlorite (Cushing 1979). When such treatment involves heating, it can be called thermo-oxidative degradation (Brogly 1978; Brenner et al. 2016).

Ultrasonic conversion

As a means to avoid possible toxic byproducts, another alternative to hypochlorite oxidation involves ultrasonic treatment (Brenner et al. 2016; Radosta et al. 2016). Such treatment was shown to cleave the amylopectin component, breaking off the branch points, while leaving the linear amylose chains relatively intact. Such a process would be expected to generate undesirable very-low-mass fragments. At a given viscosity, the starch tended to have higher penetration into the sheet (Brenner et al. 2016).

Optimized molecular mass

A topic that often gets lost in discussion of conversion of size-press starch is the fact that there is always a relationship between molecular mass and the ability of a starch product to contribute to paper strength. The effect may be negligible at relatively high molecular mass, but the effects can become large when the degree of conversion reaches excessive levels, which need to be determined by experimentation. Thus, Brandl (1984) found that excessively high temperatures during continuous cooking of starch decreased the viscosity excessively, resulting in over-usage of starch at the size press. Dobbs et al. (1993) documented a case study in which a paper mill was able to save money by backing off on the level of enzyme treatment, thus increasing the degree of polymerization of the resulting starch. Although the resulting higher viscosity meant that the operators had to reduce the solids of the size-press solution, they were still able to achieve their target paper strength levels.

Starch Cooking

There are two main ways to cook or “gelatinize” the starch grains (Leach 1965; Ai and Jane 2015), often just before their usage in papermaking. The most traditional method is a batch process, whereas most modern operations have adopted continuous or “jet” cooking procedures.

Batch cooking

Batch preparation of starch generally takes place in steam-heated vessels, as has been illustrated in Fig. 11. To ensure even mixing of the starch granules, the initial mixing with water is carried out below the softening point for the given starch product. Then the temperature is raised to the neighborhood of 90 °C and held for up to about 20 minutes, depending on the starch product. During this period, the starch granules start to swell. With continued swelling, they will become more translucent, and then starch macromolecules will start to be released. Especially when cooking potato starch, the viscosity of the mixture will go through a maximum (Leach 1965; Swinkels 1985), which is likely due to friction between a crowded population of over-swollen grains. Once the starch has been fully cooked, the structural contribution to viscosity is no longer present and viscosity reaches a plateau that depends on molecular mass and solids content. In the case of converted starch products, the viscosity levels will be much lower than the native starches, often by one or two factors of ten.

Continuous (jet) cooking

Continuous cooking of starch offers the potential advantages of smaller equipment (no need for steam-heated tanks), better opportunities for process control, and less opportunity for the build-up of dried starch at waterlines (Hiemstra 1972). On the other hand, good process control is required, with respect to temperature and pressure, to ensure optimal results (Herwig 1979). As illustrated in Fig. 12, cooking takes place under pressure within what might best be described as an enlarged pipe, following a valve in which steam is introduced to heat up the mixture. The solids content fed to the jet cooker needs to take into account the final solids target (e.g. maybe 10 to 18% in different cases), as well as any dilution needed cool the mixture to reach a target run temperature (e.g. 65 or 70 °C). Jet cooking temperatures in the range 100 to 140 °C can be used (Fanta et al. 1999; Byars 2003; Ferng et al. 2011). The time of treatment is often hard to define, due to the possibility of continued cooking after the cooked starch is diluted and released from the pressurized zone. Bradl (1984) showed that factors such as temperature and pressure need to be carefully controlled at optimum values to achieve consistently favorable results with a jet cooking system. In particular, under-cooked starch may result if the supply of steam has variable properties.