Paint analysis

Paint analysis in historic preservation entails the scientific analysis of architectural finishes, including not only paints but also metallic finishes and clear and translucent finishes used on historic buildings. The primary purpose of such analysis is to determine the color of the finish used at a particular time in the building's history, usually the original construction, but not always. Secondary purposes include determination of ingredients such as media (water, oil, latex, etc.) and pigments (organic pigments, inorganic pigments, dyes, etc.). Paint analysis is also used at times as a dating technique for various building elements.

Typical problems encountered in historic paint analysis include such things as paint loss, surface deterioration, newer materials, substrates, delamination, media and pigment deterioration, and alligatoring.

History
Historic architectural paint analysis finds its roots in the early twentieth century in the United States. The historic preservation movement began in 1849 with the preservation of Mount Vernon, the home of George Washington. Early preservationists began to realize that paints and finishes which had survived were very important but may not have been the original, or historic, finishes. Interest in historic wallpapers also developed with the interest in historic paint and color.

One of the earliest endeavors came with the restoration of Williamsburg, Virginia funded by John D. Rockfeller in the 1920s. Early investigations by simple scraping of the finishes by Susan Nash of the surviving original buildings yielded a palette that became popularly known as "Williamsburg colors". They were commercially produced and were used on innumerable Colonial Revival style buildings from the 1930s and onward.

In the 1950’s and 1960’s serious efforts at investigating original paint colors were underway at Independence National Historical Park in Philadelphia by architect Penelope Hartshorne Batcheler. Her pioneering efforts introduced, for the first time in this country, the use of a stereo microscope to more closely examine the 18th century paints at Independence Hall. Batcheler also introduced the use of the Munsell Color System for matching and referencing original paint colors. Her landmark publication, “Paint Color Research and Restoration”, was the very first publication concerning the analysis of historic architectural paints for determination of original colors. At the same time, in the United Kingdom, microscopy of paint samples was developed by Joyce Plesters of the National Gallery, London who worked mainly with easel paintings but also with samples from wall-paintings.

In the 1960s and early 1970s Morgan W. Phillips at the Society for the Preservation of New England Antiquities (SPNEA) became very involved with historic paint and color analysis, specifically at the Harrison Gray Otis House in Boston. Phillip’s definitive research and publications explained how oil paints discolor as they age. At the same time, E. Blaine Cliver, Historical Architect, who originally worked with Batcheler at the National Park Service (NPS) in Philadelphia, then with the National Trust for Historic Preservation in Washington, DC and later at the Northeast Regional Office of the NPS, became involved with historic paint analysis, especially in a laboratory in Building 28 of the former Boston Navy Yard.

In the early 1970s, as the field of historic preservation expanded, Frank S. Welsh joined the NPS in Philadelphia and began his research and study of historic paints with Penelope Batcheler, where he introduced the use of the National Bureau of Standards Color Name Charts (part of the National Institute for Standards and Testing for naming the colors matched to the Munsell Color System. As an independent historic paint color consultant one of his first major projects was Monticello, the home of Thomas Jefferson.  In the mid 1970s, Matthew J. Mosca started working for the National Trust with Blaine Cliver. Later, as a preservation consultant, Mosca researched the historic colors of Mount Vernon.

The advances in the science of paint color research by these individuals suggested that the popular Williamsburg colors had been matched to faded and aged finishes,. Therefore, during the 1980s and 1990s Colonial Williamsburg consulted with Welsh to undertake a comprehensive paint and color analysis on numerous buildings in the historic area. His research, “the first modern scientific paint analysis” there, confirmed that the Williamsburg color palette did not represent the actual historic colors. In addition he found that in many cases their early efforts had mistakenly matched later paint layers, even some that were Victorian. Welsh’s work extended also into the investigation, analysis and color evaluation of Colonial Williamsburg’s historic wallpaper collection, much of which has been published.

The interest in scientific paint analysis spread to numerous higher educational institutions across the United States. The first such graduate program was started at Columbia University with James Marston Fitch and Charles E. Peterson, then later to Boston University and the University of Pennsylvania where Frank Matero, director of the Architectural Conservation Laboratory and Professor of Architecture in UPenn's Graduate Program in Historic Preservation, continues to lead in the educational aspects of technical architectural conservation.

Many more now offer courses in the field including the College of Microscopy in Westmont, Illinois and the McCrone Research Institute in Chicago. The College of Microscopy was founded in January 2004 as a member of The McCrone Group, Inc. while McCrone Research Institute was started by Walter McCrone, one of the pioneers in the field of chemical microscopy. The College of Microscopy specializes in (1) one-week intensive courses in all areas or disciplines of microscopy, 2) an undergradte program in Applied Microscopy in conjunction with Concordia University, River Forest, Illinois, and (3) a masters degree program beginning in 2010 in Applied Microscopy. McCrone Research Institute is devoted to the teaching of microscopy, especially that which is related to the polarizing light microscope (PLM).

The College of Microscopy and McCrone Research Institute, although bearing similarities along with a history of McCrone, are two distinct and independently functioning organizations.

Journals of the Association for Preservation Technology International (APTI) and of the American Institute for Conservation, (AIC), and The Microscope serve as the primary sources for the publication of scholarly articles concerning historic paint, pigment, wallpaper, fiber and color analyses. McCrone, Phillips, Welsh and Matero have authored the majority of the articles in these journals. Others include Downs, Alderson and Perrault. Many, except those written in The Microscope, are available through JSTOR. The principle multi-author book on the subject is Paint in America.

In addition, "ModernMicroscopy.com", a free peer reviewed journal published by The McCrone Group, Inc. and the College of Microscopy appears online and contains articles of interest in many varied disciplines including art conservation and historic preservation.

Sample Collection
Historically, paint analysis was done on site by carefully removing later paint layers to reveal a sequence of finishes down to the substrate. This was the methodology employed during the early restoration of finishes at Historic Williamsburg. Although this method is employed by a few practitioners, it is not common because of its inherent problems of misinterpretation and failure to address issues such as paint ageing and discoloration.

Because finishes analysis is performed under laboratory conditions samples are collected in the field for later analysis and can be collected by the analyst or by his client who then ships them to him. They are typically collected in one of two types of packages. Manila coin envelopes are highly recommended. They have large flaps which should remain unsealed. There is virtually no possibility of the sample migrating from such an envelope. The other possibility is plastic resealable (Ziploc) bags which can be opened and reclosed at will. The only drawback to this type of package is that labeling can be difficult. Under no circumstances should paper letter envelopes, sealed or unsealed, be used. If they are sent in a sealed state they have no further value once they are opened. If they are sent in an unsealed state the sample readily migrates from the envelope as the flap is inadequate to contain the contents.

The samples are labeled during the collection process. Typical information includes the sample number, building name, building location, name of collector, date of collection, and specific data regarding the actual location of the sample. This can be written on the face of the manila envelope or, in the case of the plastic bag, can be written on it using appropriate pens or written on paper and included inside the bag.

Collection procedures vary primarily depending on the type of substrate encountered. In order of typical frequency, substrates include wood, plaster materials, wall coverings, hard masonry materials, and metals. The samples are collected using a sharp metal blade such as a scalpel or XActo knife. If the latter is used, a curved blade similar to that of a scalpel works best.

Sample size is relatively insignificant compared to quality. Although actual parts of a building are submitted for analysis, the sample need not be large at all as it is viewed through a microscope. In this case, size does not matter. What is needed is a sample with all of its paint layers well adhered to each other and to their respective substrates.

There are cases in which the paint simply refuses to adhere to the substrate. Typically this happens with wood elements which were originally primed with varnish. In these cases the samples should be collected without the substrate and if varnish was used, it will appear under the microscope, confirming the original prime coat.

Most wood elements are milled trimwork. For these one should find areas with an apparent heavy paint buildup. Weathered wood, which many think to be original, has lost its historic finishes and, invariably, proves to be worthless. Areas that are relatively protected from weathering usually retain the most complete sets of finishes. Typically vertical surfaces are better than horizontal surfaces. Worst surfaces are areas such as window sills. The sample is gently cut with the grain and pried loose. A broken surface to the cross section of the sample is best. If the sample is cut or sawn the paint layers become blended together. Typically, the wood splinter does not easily stop as the grain goes farther into the wood. If one finds this happening, then a countercut perpendicular to the grain can be made and the splinter snapped off at the point. For those wood surfaces without an edge, sample techniques similar to those of plaster, as described below, can be used.

Except for plaster moldings, plaster surfaces tend to be flat. Samples from plaster moldings can be collected using techniques similar to those described for wood edges above. For flat areas the blade can be used to create a shallow crater, making sure that, if possible, all of the paint is well adhered to the plaster. Again, the goal is to reveal a complete set of layers with rough, broken edges. Sometimes a thin, small, rectangular piece is cut from the plaster surface.

For hard masonry surfaces such as brick, stone, and concrete the same principles apply, although execution can be extremely difficult, if not impossible. In a worst case scenario, the paint can be removed to the surface of the substrate with care being taken to take samples with complete, intact, sets of layers.

For metal surfaces there is no reasonable means of collecting the sample with its substrate. If the metal piece is small it can be removed and submitted for analysis. If not, the sample should be removed as gently as possible making every attempt to remove an intact sample. Too frequently, however, this proves to be impossible and a scraping technique must be employed. Although this results in fragments and dust, even this type of sample typically can be analyzed with positive results.

A fully illustrated publication on the investigation and sampling of historic paints and wallpapers is the PaintpamphletTM written by Frank S. Welsh. First published in 1977, the fourth edition is currently available from Welsh Color and Conservation, Inc.

Laboratory Analysis
Although there are many tests which can be undertaken on paint in a laboratory setting, only basic analysis will be discussed here. The primary purposes of basic analysis are to determine historic finishes and to determine principle componenst such as media or basic pigments.

There are two methodologies in practice at present in the preparation of paint samples for microscopic analysis. The first, which is derived from the medical world, is to treat the sample as a specimen and set it into a fixed position in a permanent medium such as paraffin. The specimen is then ground to a flat finish, providing a horizontal surface for viewing under a microscope. The second is to leave the samples in a loose condition with their broken surfaces which then can be manipulated under the microscope to permit a variety of views of the layers. The primary disadvantage to the first method is that the grinding process tends to blur layers together, especially layers of similar or identical colors. It also provide only a single, fixed point of viewing. The second method lacks these disadvantages, although skill and experience is required to manipulate the samples effectively.

Following preparation of the samples, they are typically viewed under an optical microscope using either natural north light or polarized artificial light simulating natural north light. North light is essential in order to render the colors accurately without the effects of the yellow spectrum of direct sunlight.

Each individual layer is identified and, typically, matched to the Munsell System of Color. The Munsell System of Color is a scientific system in which colors have been ranged into a color fan based upon three attributes: hue or color, the chroma or color saturation, and the value or neutral lightness or darkness. Unlike color systems developed by paint manufacturers, the Munsell system provides an unchanging standard of reference which is unaffected by the marketplace and changing tastes in colors.

The hue notation, the color, indicates the relation of the sample to a visually equally spaced scale of 100 hues. There are 10 major hues, five principal and five intermediate within this scale. The hues are identified by initials indicating the central member of the group: red R, yellow-red YR, yellow Y, yellow-green YG, green G, blue-green BG, blue B, purple-blue PB, purple P, and red-purple R. The hues in each group are identified by the numbers 1 to 10. The most purplish of the red hues, 1 on the scale of 100, is designated as 1R, the most yellowish as 10R, and the central hue as 5R. The hue 10R can also be expressed as 10, 5Y as 25, and so forth if a notation of the hue as a number is desired.

Chroma indicates the degree of departure of a given hue from the neutral gray axis of the same value. It is the strength of saturation of color from neutral gray, written /0 to /14 or further for maximum color saturation.

Value, or lightness, makes up the neutral gray axis of the color wheel, ranging from black, number 1, to white at the top of the axis, number 10. A visual value can be approximated by the help of the neutral gray chips of the Rock or Soil Color chart with ten intervals. The color parameters can be expressed with figures semi-quantitatively as: hue, value/chroma (H, V/C). The color “medium red” should serve as an example for presentation with the three color attributes, 5R 5.5/6. This means that 5R is located in the middle of the red hue, 5.5 is the lightness of Munsell value near the middle between light and dark, and 6 is the degree of the Munsell chroma, or the color saturation, which is about in the middle of the saturation scale.

Report Preparation
Following the microscopic investigation a report is written. There are a variety of report types. Some analysts simply summarize their findings and provide little or no discussion of the individual samples and provide only their conclusions regarding historic finishes. At the other end of the spectrum are those who provide not only a discussion of each, individual sample, but also Munsell colors for each layer of each sample. This enables the clients to reach their own conclusions. In many cases the clients have access to additional information such as the history of the building and its maintenance which is not available to the finishes analyst.

Some analysts also provide photomicroscopy as part of their reports. Photomicroscopy is the photography of the samples through the microscope. Its advantage is to illustrate the findings of the report. Disadvantages include color distortions created by the light source and photographic dyes and also the tendency for some clients to match the colors observed in the photographs, rather than those provided in the report itself.

Following the basic report, further research may be required. Typical research includes the following:
 * Provision of color sheets of selected colors from the Munsell Color System.
 * Matching of the Munsell colors to a paint manufacturer's system.
 * Pigment testing, such as for lead content.
 * Media testing, such as for linseed oil or latex.
 * Testing of clear or translucent finishes such as varnishes and shellacs.
 * Further investigation of decorative painting such as graining (imitating wood), marbelizing (imitating marble), stencilling, or murals.
 * Consultation in the replication of historic finishes.