[Student Projects, Paper Making, Pigments and Binders]

Watercolor Paint and Hand-made Paper

photograph of paper samples after paint tests complete

Basic Information
Procedural Information
Results and Observations
Concluding Remarks

The Basics


A painting has three components: a "support" - the wood, canvas, paper, etc., on which the paint is applied; a "pigment" - the agent of color in the paint; and a "binder" - the medium which adheres the pigment to the support. A painter may vary any or all of these factors according to the results he or she hopes to produce. This project studies the effects of varying one of the above three components/factors. The project revolves around pigments in a gum arabic binder (watercolors). The variable will be the paper which serves as a support.


This project studies the relationship between watercolor paints and hand-made paper.



Layman's Terms

Test of how paper samples absorb various watercolor paints and effect on apperance of paint


At the start of this project, I expected that changes in the paper would not affect the appearance of the watercolors. However, I was open to the possibility, especially since changes in appearnce would make my project more interesting.


Making the watercolors

Making the paper

Jamie MacGregor had boxes of beaten pulp, with pulp type and beating time documented. I used two pulps with two beating times each, as follows (abbreviations for further notice in parentheses): cotton fiber beaten 10 minutes (C10), cotton fiber beaten 20 minutes (C20), linen fiber beaten 11 minutes (L11), and linen fiber beaten 16 minutes (L16). I also experimented with an internal sizing agent. In total, I pulled two sheets of each pulp/beating time combination, one with sizing and one without, resulting in 8 sheets overall.

Variation/Addition in Procedure for Sizing

Before being able to pull the sheets used in the test, I had to determine the proper number of drops of sizing to use per sheet. Since I was trying to control the amount of sizing for a single sheet, I couldn't just add the sizing to the entire batch of pulp. Using ratios of pulp poundage to mililiters of sizing, given to me by Dr. Bordley from a previous experiment, calculated an approximate number of drops to use. I then pulled five sheets of a single pulp type/beating time combination, varying the amount of sizing around my calculation. I varied the number of drops from 5 to 10, 15, 20, and 25 (15 being the original approximation). After pulling and ironing the sheets, I then spread on paints of each pigment to observe the effects of the sizing. The sheets with 5 and 10 drops absorbed the watercolors too quickly and the paper itself rolled away with the brush strokes. For the sheets with 20 and 25 drops of sizing, the paints stood on the surface of the sheets before finally being absorbed. The paints did not spread well or evenly. I determined that the original approximation of 15 drops of sizing per sheet was correct. The following list includes the steps added to the papermaking procedure to add sizing. This method is for internal sizing (separate from external, which is added after the sheet has been pulled).

Testing: one, two...

Materials needed: paintbrush, small cup of water, watercolor paints, hand-made paper samples, pencil (for labelling), colorimeter.

Observations and Data

The Paint

The ultramarine and yellow ochre made good "cakes" of paint, as can be seen in the photographs. The green earth made a cake of paint, but it was not smooth like the ultramarine and yellow ochre. The red earth would not cohere enough to make a nice cake of paint, but by pushing the mixture into the corner of the weighing boat I was able to make a thick enough sample. When painting with these pigments, the ultramarine was the loudest. The green earth seemed a little washed out. L*a*b* values show these aspects scientifically.

Also viewed paints through microscope.
View microscope pictures

The Paper

In the process of making the paper, the pulp that was beaten longer took longer to drain. This is a result of fibrillation. As pulp is beaten, the fibers are severed and nicked. They then absorb water more easily because the cellulose inside is hydrophilic (water-loving). The fibers also attach more closely to each other. The water thus takes longer to drain out.

Physically, the linen paper was smoother and absorbed the paints more evenly. All the sheets made with sizing seemed "crinkly" and stiff compared to the same pulp/beating time combination without sizing. The "nicest" paper for painting on seemed to be the linen beat for 11 minutes without sizing. Its texture and absorption of the paintmade the paints spread smoothly and produced good color values.

The Measurements

DISTANCE (cm) C10, no size C10, size C20, no size C20, size L11, no size L11, size L16, no size L16, size
Ultramarine 5.8 6.0 7.0 7.1 6.8 7.6 5.5 7.0
Red earth 6.1 5.0 7.4 6.35 7.4 6.05 5.8 5.95
Green Earth 6.4 5.85 7.5 6.85 8.25 5.0 6.4 5.75
Yellow Ochre 6.3 5.65 7.6 5.75 7.55 5.5 6.5 6.45

All the pigments besides green earth spread farther (on average) on the cotton paper than on the linen. On average, all the paints except ultramarine spread farther on the sheets without sizing. Information like this is important to painters. If a paint spreads farther on one kind of paper than on another, then the painter will know that on the first sample a specified area can be covered with less paint. The painter is then able to conserve paint by using the paper which allows him/her to cover the same space with less paint.

L*a*b* C10, no size C10, size C20, no size C20, size L11, no size L11, size L16, no size L16, size
Ultra- marine L=48.19 a=+6.42 b=-16.78 51.12 +5.63 -15.49 44.95 +8.68 -21.26 48.91 +6.42 -16.41 48.36 +8.28 -21.25 48.69 +6.55 -18.47 49.28 +6.57 -16.27 47.42 +8.38 -20.11
Red Earth 56.21 +5.16 +5.79 58.63 +6.33 +8.57 55.08 +8.84 +9.22 53.85 +8.77 +7.85 53.62 +8.37 +8.19 56.39 +9.65 +11.40 57.86 +7.84 +9.88 56.48 +7.76 +10.22
Green Earth 56.43 -3.75 +5.53 59.18 -4.38 +8.47 62.60 -3.33 +5.67 65.44 -3.22 +6.51 55.29 -5.06 +8.10 58.50 -4.54 +8.82 60.01 -4.75 +8.22 56.51 -4.11 +7.68
Yellow Ochre 58.77 +3.11 +15.53 63.19 +0.95 +13.34 60.54 +2.85 +16.89 63.93 +0.13 +12.13 58.70 +2.67 +14.36 63.02 +0.55 +13.91 57.59 +3.85 +17.52 63.47 +1.28 +17.77

The L*a*b* color space is a graph with 3 axes. The horizontal axis is named "a." Its values range from -60 to +60. The negative end (from 0 to -60) represents green, while the positive end (0 to +60) represents red. The larger the number (disregarding the +/-), the more intense the color represented by the + or -. The vertical axis is named "b." Like the "a" axis, the "b" ranges from -60 to +60. The negative values are blue, and the positive values are yellow. Again, the larger the number without regard to sign, the more intense the color named by the + or -. The third axis is named "L," and this one extends through the intersection of "a" and "b." The "L" values can range from 0 through 100, and the scale represents lightness. 0 is black, and 100 is white. The L*a*b* color space offers a way to describe the appearance of a color scientifically, as numerically it leaves room for well over 1.4 million colors, each with different L*a*b* coordinates.

The L*a*b* coordinates for a given pigment cam be important to a painter. Raising and lowering the values results in new shades and hues. A painter needs to know how paint typically appears on a certain kind of paper. L*a*b* is one way to find out. For instance, my colorimetry data shows that on linen papers with sizing, the yellow ochre paints were, on average, darker than on linen paper without sizing. This would be important for a painter to know before using than particular paper.


What did I learn from this experiment?

My realizations arising from this project are first about how many factors can affect the physical characteristics of a piece of paper. The type of fiber used, the length of beating time of the fibers, the amout of sizing used, and the nature of sizing used are only some of the variables involved. I knew that many kinds of paper could be and are manufactured, but I was unaware of the variables prior to this project. I have since learned from Dr. Kirven that several institutes exist for the sole purpose of studying paper. Click here to visit the
Pulp and Paper Centre at the University of Torornto. I had expected that the variances in the paper would not affect the appearance of the watercolors. However, fiber type, beating time, and sizing each made an impact on paint appearance. I learned that watercolorists have a lot to worry about and must pay attention to the paper they use for a support. I also learned how much chemistry is involved in the seemingly artistic areas of papermaking and watercolors, because my project included so many subjects covered in the "Chemistry and Art" class. This experiment drew on everything from the chemistry of handmade paper (including the cellulose structure of the fibers and similar structure of the sizing) to the use of pigments and binders to L*a*b* color space theory.
So, while my project didn't find anything scientifically exciting (like discovering a new element), it did establish an important link between the chemistry of paper and paint making and the art of watercolor. It truly was a "Chemistry and Art" project.

Lonsdale MacFarland Green, 1997.