The irony is that neither starch nor gelatin is water repellant! The molecular components of starch—amylose (fig. 1) and amylopectin (fig. 2)—greatly resemble cellulose (fig. 3), the principal component of papermaking fibers, and are every bit as hydrophilic (water loving) as cellulose. Gelatin, though possessing a different structure derived from protein, is also hydrophilic. Amylose and cellulose are both chains of repeating anhydroglucose sugar units, differing only in how they are linked together. All solid lines in these structures represent the strongest bond in chemistry—the covalent bond—which holds the amylose and cellulose backbones together and will become important when we address paper degradation by acids. There are two approaches to impart water repellency in paper. The first is to physically block or constrict the microscopic channels or pores through which water migrates. This is what starch- and gelatin-based surface sizes do.6 The second approach is to incorporate a hydrophobic (water fearing) substance into the paper. We know this as internal sizing, developed by the paper industry to size paper during sheet formation. That industrial internal sizing dates only to about 1807 when Moritz Friedrich Illig invented rosin–alum sizing7 shows us the technical challenges of internal sizing using hydrophobic substances. Illig's approach was to use rosinous acids that were first tapped from trees, but were later obtained from wood-pulping operations.8 While there are several approaches to rosin–alum sizing, the object is to fix rosinous acids to the paper fibers using aluminum ions introduced to the stock as potash alum or papermakers' alum.9 (See fig. 4.) This process has much in common with fixing acid dyes, with the exception that you are linking a hydrophobic rather than a colored molecule. Rosinous acid molecules, though mostly hydrophobic, each have a hydrophilic end that connects via the aluminum ion onto the fiber, orienting the hydrophobic end to the fiber surface. Rosinous acids include abietic acid (shown in fig. 4), levopimaric, and fumaropimaric acids. The introduction of these hydrophobic chemicals— at the speed of the paper machine, into the furnish and retained evenly throughout the finished sheet—represents a major technical achievement in industrial papermaking. Alas, the problem with rosin–alum sizing is that it both depends on—and creates—an acidic environment. When aluminum ions are introduced as any salt such as potash alum or papermakers' alum to a pH-neutral, aqueous environment such as a papermaking stock, the stock becomes acidic.10 They do this by grabbing water molecules, splitting a few of them apart, holding onto the OH- but releasing the H+ into solution, bringing the pH down. If you try to compensate by adding a base to the stock, the aluminum precipitates prematurely as another salt, never making it into the sheet. Hence, industrial papermakers who are using alum keep the stock at about pH 4.2 to 5.2 (adding a second source of acid if necessary) so that it bonds with rosins and flocculates onto the fibers at just the right time.11 In the finished sheet, the linkage of rosinous acids to the papermaking fibers, via an aluminum ion (shown with dashed lines in fig. 4), is a distinct form of bonding called dative bonding that is weaker than the covalent bonding described earlier. The significance of this weaker bonding is not that the sizing becomes unstable, but that the aluminum ion is figure 4 Skeletal structure of rosin–alum sizing showing cellulose backbone (top), abietic acid (bottom), and aluminum ion linkage (middle). Dashed line in cellulose is a hydrogen bond and other dashed lines are dative bonds to aluminum ion. In addition to cellulose and abietic acid, representative ligands are shown including water and hydroxide. sufficiently free to interact with other molecules in a way that is classified according to Gilbert N. Lewis (1875–1946), in his widely applied alternative definition of acids and bases, as being acidic (creating a third source of acid). So why exactly is acidity bad for paper? Acidic compounds within paper can have a destabilizing effect on painted, printed, or written media on the paper to an extent that will depend on media composition. Acidic compounds also have adverse effects on adjacent materials, or materials separated by a distance if the acids are volatile. The largest problem with acidity in paper, however, is the corrosive effect it has on the paper itself due to splitting the cellulose chain (fig. 3), breaking the covalent bonds that hold the chain together. Attributing paper brittleness to this molecular event has been established through several analytical methods, but part of the story is told by holding a magnifying glass to the torn edge of both an aged, acidic sheet of paper and a healthy one. While the fibers in the healthy sheet will jut from the edge, the fibers in the acidic paper appear broken, evidence that the degradation is within the fibers themselves.12 The broken cellulose chain is itself a source of acidity (thus a fourth source of acid for rosin–alum sized papers). Two new chain ends are created by the break, which can oxidize (or rather ends facing right in fig. 3 can oxidize) to organic acids, which in turn catalyze further chain breaks and so on in a spiraling effect. Acknowledging problems associated with acidity from using rosin–alum sizing, the paper industry began in the 1950s to commercialize internal sizing agents for pH neutral (or mildly alkaline) sizing.13 Figure 5 shows the mechanism of anchoring the first class of these chemicals—alkyl ketene dimers (AKD). Both Aquapel and Hercon products, which are brands known to many paper technologists and artisans, are AKD sizing systems, all of which feature a group that reacts to form covalent bonds onto the cellulose backbone, specifically an ester group linkage. While the term reactive sizing derives from this chemistry, it is more useful in describing the evolution of industrial alkaline sizing rather than describing the stability of this sizing. As we have seen, rosin–alum sizing is bound to the cellulose only via dative bonds, and surface sizes do not even have that. Shortly after the first development of AKD, a second class of industrial internal sizes evolved, alkenyl succinic anhydrides (ASA). The mechanism of ASA anchoring is shown in figure 6, which also creates an ester linkage with the cellulose backbone. ASA can have certain advantages in industrial papermaking, but are more reactive than AKD and therefore should be used promptly. In the presence of water, the sizing can be prematurely "quenched," the product of which is a double organic acid that will be neutralized in an alkaline environment, but which will not achieve sizing. When selecting a sizing for particular applications, there are three main tests that assess the overall sizing performance of internal or surfaces sizes—whether hydrophilic or hydrophobic—in handmade or machine-made papers. These tests— the angle of contact method, the Cobb test, and the Hercules test—assess sizing in very different ways.14 The angle of contact method (fig. 7) assesses the surface wettability of paper versus time by placing a water droplet on the sheet and carefully measuring the angle that the edge of the droplet forms with the paper surface. In highly water-repellant papers, the droplet will stay orbiculate (a), much like a bead of mercury, but in more bibulous15 papers, it will flatten out considerably (b). The Cobb test (fig. 8) measures how much of a measured volume of water can be absorbed by the paper in a specified period of time, where the water is confined to a definite surface area of the paper by a brass ring clamped above it. The Hercules test (fig. 9) measures the time it takes for a photometer to observe a specified loss of light reflectance on the bottom surface of a paper sheet after a standard ink is poured into a bottomless cylinder resting above it. Of the three test methods, the Cobb test is perhaps most useful for handmade papers because it assesses sizing as a bulk property of a sheet rather than at a specific point, which better accounts for the variation inherent in handmade papers. Porosity is another important criterion to assess surface sizing. Researchers currently use the Gurley and Sheffield methods to extrapolate paper porosity from the rate of air penetration through a sheet, given specific experimental conditions. 16 I expect we will soon see digital-image analysis applied to porosity studies as it has been to analyze archaeological ceramics.17 Sizing can also influence optical properties of paper such as brightness and opacity, although these properties are mainly dependent on a paper's structure.18 Opacity is attributable to free surfaces within the sheet that are available to reflect light, with about seven fiber layers often sufficient to make an opaque sheet. Both rosin–alum sizing and its pH-neutral successors closely conform to the fiber surface, bonding at the molecular level. Therefore, internally sized papers will have much the same opacity as their unsized counterparts. Because these chemicals do not absorb light in the visible spectrum, they will also not darken the paper. Surface sizes are a different matter. Given the objective to limit porosity, surface sizes begin to fill the sheet, reducing free surface area. A similar effect is seen in papers made from highly beaten fibers, which readily conform to one another, limiting reflecting surfaces. While gelatin may have a yellowish cast, surface sizes should generally not absorb visible light; the darkening effect is rather due to the reduction in the ability of the sheet to scatter light, letting it pass through the sheet. While I perceive this subtle effect is expected in Western papermaking using gelatin sizing, which again emulates the translucency of parchment,19 there has been some uncertainty about this effect in traditional Middle Eastern papermaking. The most in-depth descriptions of the process,20 however, acknowledge that brightening is due to additional components in the sizing such as starch powder that add light-reflecting surfaces to the sheet. Hand papermakers today have some choices of sizing agents. In a review of catalogs from hand papermaking suppliers, it appears that most offer a house-brand AKD sizing system (internal) or gelatin (surface) or both. No ASA sizes seem to be available, no doubt due to the impracticality of using a product that chemically degrades to a significant extent within minutes after its emulsification. An AKD formulation that has been thoroughly tested should offer a permanent and stable sizing. Things to look out for are impermanent groups as side chains, rendered –R1 and –R2 in figure 5, and the identity of other ingredients such as emulsifiers and stabilizers. AKD, like ASA, can react with water to form a non-sizing organic acid, but this happens slowly and the acid is neutralized in an alkaline environment. Any recommendation for permanent and stable surface sizes for hand papermaking must draw on the work supporting resizing of cultural heritage papers in conservation. In 1986, a survey of conservation practice in the United States showed a slight preference for cellulose ethers over gelatin for this purpose.21 Most recently, however, there has been considerably more work reported on gelatin resizing over any other method, certainly within the United States. Cellulose ethers have the advantage figure 7 Illustration of the principal of assessing surface wettability of paper via the angle of contact method. In heavily sized papers (a) the angle the droplet makes with the paper (θa) is relatively acute. In less heavily sized papers (b) the angle is more obtuse (θb). of being composed of much the same material as paper—cellulose. They have been extensively reviewed as conservation materials and have clear applications in paper-based collections.22 Gelatins offer the advantage of being a natural product with centuries of history in papermaking. And unlike other sizing agents, gelatins can actually buffer paper pH, thereby improving its permanence and stability.23 If you have ever measured pH using an electrode, you have calibrated the electrode using a special kind of buffer. Gelatins help maintain paper pH via the same principle, and the great permanence of many Western papers prior to industrial internal sizing must be due in part to this effect. Call me a traditionalist, but I think technologies that have this kind of track record in cultural heritage materials are great candidates for use in our work today. ___________ notes 1. Sukey Hughes, Washi, the World of Japanese Paper (Tokyo: Kodansha International, 1978), 40. Timothy Barrett with Winifred Lutz, Japanese Papermaking: Traditions, Tools and Techniques (New York: Weatherhill, 1983), 234, 285. Aimee Lee, Hanji Unfurled: One Journey into Korean Papermaking (Ann Arbor, MI: The Legacy Press, 2012), 30, 162. 2. Helen Loveday, Islamic Paper: A Study of the Ancient Craft (London: Don Baker Memorial Fund, 2001), 44. Joseph von Karabacek, Das Arabische Papier (Vienna, 1887), trans. Don Baker and Suzy Dittmar as Arab Paper (London: Archetype, 2001), 42–43. Jonathan M. Bloom, Paper Before Print: The History and Impact of Paper in the Islamic World (New Haven: Yale University Press, 2001), 69, 73. 3. Richard L. Hills, "Early Italian Papermaking, A Crucial Technical Revolution," IPH Congress Book (International Association of Paper Historians, 1992), 9, 37–45. 4. Loveday, Islamic Paper: A Study of the Ancient Craft, 42. 5. John D. Peel, Paper Science and Paper Manufacture (Vancouver: Angus Wilde Publications, 1999), 217. 6. Christopher J. Biermann, Handbook of Pulping and Papermaking (San Diego: Academic Press, an Imprint of Elsevier, 1996), 245. Gerhard Banik and Irene Bruckle, Paper and Water: A Guide for Conservators (Amsterdam: Butterworth- Heinemann), 152. 7. Dard Hunter, Papermaking: The History and Technique of an Ancient Craft (New York: Dover Publications, 1974), 523. 8. Gary A. Smook, Handbook for Pulp & Paper Technologists (Vancouver: Angus Wilde Publications, 2003), 221. 9. For more on rosin–alum sizing for a conservation audience, see Martin A. Hubbe "Acidic and Alkaline Sizings for Printing, Writing, and Drawing Papers," The Book and Paper Group Annual 21 (2002): 139–51. 10. John C. Roberts, The Chemistry of Paper (Cambridge: The Royal Society of Chemistry, 1996), 125. 11. Peel, Paper Science and Paper Manufacture, 81. 12. John W. Baty, et al. "Deacidification for the conservation and preservation of paper-based works: A review," BioResources vol. 5, no. 3 (2010): 1955–2023. 13. J. W. Davis, W. H. Roberson, C. A. Weisgerber, "A New Sizing Agent for Paper—Alkylketene Dimers," TAPPI Journal vol. 39, no. 12 (1956): 21–23. 14. See TAPPI (Technical Association of the Pulp and Paper Industry), www.tappi.org, for descriptions of test methods T 458 cm-14 "Surface wettability of paper (angle of contact method);" T 441 om-13 "Water absorptiveness of sized (non-bibulous) paper, paperboard, and corrugated fiberboard (Cobb test);" and T 530 om-12 "Size test for paper by ink resistance (Hercules-type method)." 15. Bibulous is the Technical Association of the Pulp and Paper Industry's term for unsized, intentionally absorptive papers. Amusingly, a Bing or Google search of the term yields a single definition at the top of the page, "excessively fond of drinking alcohol." 16. See TAPPI (Technical Association of the Pulp and Paper Industry), www.tappi .org, for descriptions of test methods T 460 om-02 "Air resistance of paper (Gurley method)" and T 547 om-12 "Air permeance of paper and paperboard (Sheffield method)." 17. Chandra L. Reedy, Jenifer Anderson, and Terry J. Reedy, "Quantitative Porosity Studies of Archaeological Ceramics by Petrographic Image Analysis," MRS Proceedings 1656 (January 2014): mrsf13–1656 – pp07–03. Doi:10.1557/opl.2014.711. 18. Mikko Alava and Kaarlo Niskanen, "The Physics of Paper," Reports on Progress in Physics vol. 69, no. 3 (2006): 669–723. Doi:10.1088/0034-4885/69/3/R03. 19. Ibid., 3. 20. Loveday, Islamic Paper: A Study of the Ancient Craft, 47. Von Karabacek, Das Arabische Papier, trans. as Arab Paper, 47. 21. Walter Henry, "Resizing Following Aqueous Treatment: Current American Practice," Book and Paper Group Annual 5 (1986), http://cool.conservation-us .org/coolaic/sg/bpg/annual/v05/bp05-12.html. 22. Robert L. Feller and Myron H. Wilt, "Evaluation of cellulose ethers for conservation," Research in Conservation 3 (Los Angeles: The Getty Conservation Institute, 1990). 23. John Baty and Timothy Barrett, "Gelatin size as a pH and moisture content buffer in paper," Journal of the American Institute for Conservation vol. 46, no. 2 (2007): 105–121. 24. Gary A. Smook, Handbook for Pulp & Paper Technologists, 286. Roberts, The Chemistry of Paper, 147. 25. Ibid., 24. 26. Peel, Paper Science and Paper Manufacture, 82. Roberts, The Chemistry of Paper, 129. 27. Ibid., 26.