JAIC 1985, Volume 24, Number 2, Article 5 (pp. 104 to 109)
JAIC online
Journal of the American Institute for Conservation
JAIC 1985, Volume 24, Number 2, Article 5 (pp. 104 to 109)

THE DETECTION OF MORDANTS BY ENERGY DISPERSIVE X-RAY SPECTROMETRY

N. Indictor, R.J. Koestler, & R. Sheryll



3 RESULTS AND DISCUSSION

WIDE COLOR VARIATIONS WERE ACHIEVED by the use of different mordants and additives. The same dye, cochineal, was used for each batch but with a different mordant and/or additive. In the absence of mordant or additive (sample 13) the wool fibers are nearly free of color. Each of the numbered samples was visually distinct. Visual examination of the fibers of each sample other than sample number 13 showed color well dispersed over the fibers. Table I shows the approximate sample size used for the SEM-EDS analyses and the elements observed in EDS scans of the carbon coated samples. At least three scans were performed for each sample. The table is divided into two parts: elements observed during each scan and elements observed in some scans. The frequency with which the elements were observed is also given. As expected for wool samples sulfur is prominently present in each scan. Elements other than those from the mordant may be picked up by the wool from the wash water or the non-uniform wool-mordant-dye-additive system.

Table I WOOL. Results of EDS Scans

Table II summarizes the mordants identified by EDS analysis. Samples 13 and 15 gave such small weight percentages (approx. 1%) of aluminum and iron in some EDS scans that they were reported as probably absent. Samples 20, 24, 25, 27 and 28A gave trace analyses for aluminum or iron (about 1–2%) in all EDS scans and are reported as possibly present. Samples 23A and 26 had aluminum (4–6%); samples 14 and 24 had copper (16–70%); samples 16 and 27 had iron (9–19%); samples 17 and 28A had tin (about 70%); samples 25 and 26 had chromium (17–32%). The decision to report the presence or absence of elements was based on the peak heights of these elements relative to that of sulfur, the element persistently present in all of the scans. “Yes” values were reported when EDS scans showed the mordant element to be always present in amounts greater than 2% relative to sulfur. Variation in weight percent relative to sulfur was often greater than 100% from scan to scan emphasising the already noted difficulty in using EDS analysis for quantitative assay. 1 The percentages indicated above represent only the weight percents of elements above the atomic weight of fluorine detected in an individual scan and not the weight percent present in the sample.

Table II Detection of Metal Elements on Cochineal Dyed Wool Samples by EDS

The reporting of absolute weight percents for the elements under discussion would require: preparation of known samples with controlled quantities of additives uniformily distributed on the sample; or an independant elemental microanalysis of the samples of levels greater than the 10% obtained by AA (see below). It would also be necessary to perform replicate scans to account for non-uniformity of the surface and non-uniform attachment of the additives.

Table III shows the mordants and additives used in the preparation of the samples of Table I. It is seen that all the mordants actually used were unambiguously detected (cf. Table II). The presence of traces of aluminum and iron in some of the EDS scans suggests a small problem. Continued experience and information concerning the context of samples will help the analyst to know when to disregard trace element values. In this connection the assistance of the curator or conservator cannot be overestimated. It should be noted that sample 26 had a double mordant (aluminum and chromium) and each was detected unambiguously. No difficulty at all was encountered in the detection of Sn or Cr (cf. ref. 1).

Table III Mordants and Additives Used for Cochineal Dyed Wool Samples

Table IV provides data obtained independently by AA analyses for the elements present in the mordant and detected by EDS scans. The data of column 3 is calculated from the recipes and assumes that all mordant used was incorporated into the yarn. As expected from our previous study1 the quantity of mordant found in the dyed fabric is substantially less than the amount actually used except for the tin mordanted samples. The AA results suggest that a larger amount of tin compound was used than was reported in the experimental section. The data of Table IV also suggests that the process of dying after mordanting may produce a greater retention of mordant in the fiber than simply applying mordant. Previous studies showed retention of mordants to be approximately 10%.1 The entries under % Element by AA are replicates (since each mordant is applied according to the same recipe each time it is applied). It is seen that the values are excellent replicates in some cases but in other cases differences considerably exceed the experimental error of the analysis. One reason for the differences may be non-uniform sample preparation. The procedures for mordanting and dying maintained constant proportions among the ingredients but quantities of water were not measured in the mordant or dye baths or for the washing. The additives and impurities, furthermore, may have contributed differently to the partitioning of mordant between the textile and the aqueous phase. Another reason for the differences may be non-uniformity of the samples. Wool structure may vary considerably microscopically and macroscopically.5 Animal breed, treatment, processing, aging, differences in exposure to the environment, etc., all contribute to structural, mechanical and chemical variations in wool and its ability to interact with mordants, other additives, and impurities. Dot mapping scans of our samples indicated that the individual samples were quite uniform with respect to the metallic elements in the fiber surfaces.

Table IV Comparison of Mordant Quantities Applied to Cochineal Dyed Wool Samples with Quantities Analysed by AA

Each of the samples was examined by SEM photomicrography and the samples retained. Although not prepared with this investigation in mind, the samples examined in this study represent a good standard series since the same wool and dye were used throughout; the only difference in the samples should arise from the use of different mordants and/or additives. No clear-cut distinctions in surface appearance could be made among the various mordanted and dyed wool fibers. It is possible that surface characteristics of aged fibers provide a better indication of treatment.

The distribution of the mordant on fibers is an excellent subject for the elemental dot-mapping6 capabilities of SEM. Preliminary results indicate complete delocalization of elements without discernible pattern in the surfaces of the mordanted wool fibers. This result suggests that representative EDS results ought to be obtainable even with sample sizes as small as a single fiber.


Copyright 1985 American Institute for Conservation of Historic and Artistic Works