STATE OF CONSERVATION

The state of conservation of this work is the usual one for a cellulose composition piece. Due mainly to its contact with the plywood, which forms the frame seal, the glue that sticks the piece to the base and the exhibition of the piece, it is in an obvious state of oxidation. That oxidation has altered the colour and physical characteristics of the work. In addition, it should be said that some small tears can be observed in certain areas, which are probably due the piece being moved and the pendulous vibrations to which it may have been subjected.

PROPERTIES OF CELLULOSE / THE EFFECT OF LIGHT

The effects of sunlight on cotton are well known. Both sunlight and artificial light sources simulating sunlight have been used to study them. Agreement has only been reached on the fact that the reactions that take place are very complex and of an oxidative nature. However, their mechanism of action has not been clarified.

For light to cause a chemical change in a substance, the light rays must be absorbed and have enough energy to modify the molecular structure of the irradiated product. On the other hand, the quantity of light absorbed by a fabric depends on the texture of the weave and the yarn. In that respect, the most interesting structural parameters are the yarn's twisting, retwisting, crimping and weave. Thus, high-density heavy fabrics made from very twisted yarns are least affected by light radiation. The intensity of chemical change also depends on the frequency of the absorbed radiation. Higher frequency radiation that has a shorter wavelength causes more significant changes since there is more energy per unit of radiation.

Raw and bleached cottons differ in terms of speed of photochemical deterioration. Bleached cottons are found to deteriorate faster than raw cottons when exposed to light. Attempts have been made to explain this phenomenon in two different ways:

1. The materials associated with cellulose in raw cotton act like a filter to separate harmful radiation.
2. bleaching of cotton weakens the cellulose links, which may mean that they break more easily when cellulose is exposed to light. For untreated fabrics, ultraviolet wavelengths between 2500-4000 A are considered to be the worst in terms of their deterioration effects. Radiation in that region has a high energy content per unit of radiation, and its absorption by the cellulose causes faster deterioration over equal exposure periods than other parts of the spectrum. It is important to point out that cotton deteriorates more when exposed only to ultraviolet light than when it is exposed to the same quantity of that radiation combined with radiation in the visible region of the spectrum. That phenomenon has been linked to the ease with which ozone is formed and destroyed. The most effective wavelength for converting oxygen into ozone is 3230 A (ultraviolet band of the spectrum), and the most effective wavelength for deactivating it (O₃®O₂) is 6010 A (the orange band of the spectrum). So, the combined presence of both types of radiation would lead to the formation of ozone and then its deactivation as soon as it was formed.

The effect of sunlight on cellulose were able to be studied in greater depth after more experimental data became available, and two probable mechanisms depending on the ultraviolet region that a fabric was exposed to were postulated.

The absorption of far-ultraviolet radiation by cellulose causes a photolysis or direct division of the carbon-carbon and carbon-oxygen links. To do that, energy of 80-90 kcal/mol is required, which can only be provided by short wavelength radiation with a high energy content, such as 2537 A radiation (mercury) corresponding to far-ultraviolet. Studies on these radiations in atmospheres with and without oxygen led researchers to deduce that the effects are the same. In both cases, storage in the darkness after irradiation led to subsequent deterioration, whereas the irradiated samples' exposure to oxygen in the absence thereof can lead to the oxidation of the deterioration products. It is also known that the presence of vat colorants inhibits deterioration in an atmosphere of air and oxygen, and in an atmosphere of carbon dioxide, nitrogen and helium. However, it should be pointed out that such inhibition reduces the degree of deterioration but does not stop it. Dyeing with other colorants only inhibits deterioration to a certain degree, and the presence of water vapour inhibits deterioration until it reaches a degree of polymerisation of 1150.

The phenomenon of photosensitization presents itself with near-ultraviolet. Even though cellulose and most products derived from it are fairly transparent to radiation in the visible regions and near-ultraviolet, when cellulose products are exposed to ordinary light some deterioration is often found. In such systems, where the reactant is unable to absorb the light falling on it in appreciable quantities, a chemical modification will only occur when there is a substance capable of absorbing the light and carrying the resulting energy to the reactant molecules. This phenomenon is called photosensitization and, as far as textile materials are concerned, it is linked with three very important factors: the existence of sensitisers, oxygen and humidity.

Sensitisers include a series of colorants and pigments, like zinc oxide and titanium dioxide, which are capable of absorbing the radiation that supplies the energy to deteriorate cellulose. The light energy absorbed by the colorant is transferred to the oxygen in the surrounding atmosphere, providing activated oxygen or ozone. Then the activated oxygen reacts with the water vapour to form hydrogen peroxide.

From the studies done on the relationship between the phenomenon of photosensitization and the chemical structure and dye family of the colorants that cause it, it can be deduced that:

1. Because of their solidity, vat colorants are usually employed to dye fabrics which are subjected to the effects of the weather during normal use. Several colorants in this family increase the resistance of cellulose to deterioration. Photochemical or photosensitising activity is limited to yellow, orange, red and brown.
   
   
2. In the same way as for vat colorants, there is a correlation between the colour of sulphurous colorants and their photochemical activity. Blue and green colorants of this family are less active than yellow and brown ones, some of which cause considerable photosensitization
   
   
3. Basic colorants that are not very solid when exposed to light behave like photosensitisers, although there does not appear to be any relationship between the colour and the activity. Violet, green, yellow and orange colours make deterioration faster.
   
   
4. Even though they have been fully studied, the only direct colorant that has a sensitising action is primuline. Its activity is put down to its molecular structure since, unlike most direct colorants, it is a thiazol and not an azoic colorant. It is worth pointing out that when the amino groups of primuline separate and join to a fibre with phenol, a dye is obtained which, when exposed to light, causes little or no acceleration of photochemical activity. That effect suggests that direct colorants do not speed up photochemical activity and that the sensitising activity of the thiazol nucleus decreases when an azo group is inserted into the primuline molecule.

The sensitising effects of vat and sulphurous colorants on cotton considerably decrease when, after dyeing, the fabric is treated with a copper salt. These findings are surprising, since it is known that copper speeds up photochemical deterioration in cotton. It has been suggested that copper might prevent the formation of hydrogen peroxide in dyed fabrics exposed to light and humidity. It is assumed that the presence of copper in a 40-parts-per-million proportion is sufficient to prevent the oxidation of cellulose, and it appears that the optimum amount of copper is the one that is necessary to salify the acid groups of the colorants.
Other processes for slowing down photodeterioration are based on the use of certain inactive colorants which protect the fabric from the presence of other colorants that have demonstrated activity, though the method can only be considered useful for dyes requiring a blue or green component.

Humidity increases cellulose deterioration by photosensitization. So, the presence of water vapour not only increases deterioration in fabrics and yarns containing a photosensitiser (colorant, pigment, etc.), but also speeds up deterioration in adjacent fibres. The effect is the same for distances between the undyed fabric or woven yarn at intervals between 0.3 and 8mm. Regarding the influence of water vapour on activation, it is interesting to point out that the behaviour of cotton is very different from viscose rayon. At relative humidities of 0% and 100%, cotton loses 23% and 42% of its resistance, respectively. Under the same conditions, viscose rayon loses a constant 17% of its resistance.
Photochemical deterioration in dyed cotton may be a consequence of the combined effect of direct and indirect cellulose oxidation caused by hydrogen peroxide, with the peculiarity that radiation can weaken the glucoside links and leave them easily open to oxygen attacks. The speed of deterioration increases with humidity content, and the products produced by deterioration have oxycelullose reducing properties.

When fabrics are exposed to the inert atmospheres of carbon dioxide or nitrogen, whether or not in the presence of water, there is only some slight deterioration. But as soon as they are exposed to an oxygen atmosphere in the absence of radiation, rapid deterioration occurs. The higher the exposure temperature, the faster the deterioration.

Studies done on the photoexcitation of simple compounds that are structurally similar to cellulose have led to some findings about photosensitization. The compounds studied range from primary and secondary alcohols to glucosides, disaccharides and polysaccharides. The findings show that the deterioration process begins when the photoexcited colorant molecule captures a hydrogen atom from the cellulose molecule, on which oxygen is subsequently fixed with the formation of a peroxy radical. The reactions that follow depend on the location of the initial attack. In one case, deterioration in the peroxy radical will cause the immediate division of the cellulose chain. In the other, a carbonyl group can form inside the chain, thereby suggesting the insertion of a labile link in it, which facilitates the rupture of the chain at that point, when the cellulose is subsequently exposed to the action of alkalis during the washing process.

Analysis of colour:
For this piece, readings were taken from certain points on both types of paper used by the artist. They appear to be two different colours (both green). Readings were taken for both types of paper from areas shielded from light and areas exposed to light. The figure below shows the reading points with a letter identifying the paper (exposed to light or not) from which the data were obtained..

The diagram below shows the dispersion curves for diffuse reflectance data. These data were used to calculate data given in the table above.

1. Regarding diffuse reflectance, the colour of shielded paper A (reading 278) is a well-defined green with a very sharp dispersion curve (dominant wavelength is 506nm). However, when that paper was exposed (reading 277), the values of the reflectance curve increased in the yellow area of the spectrum (the paper is turning yellow) and the chromatic data corroborate that the dominant wavelength is 571nm as opposed to 506nm for the same paper that had not turned yellow.
   
   
2. Regarding paper B, the curves for shielded areas show a dispersion that is very different from the one for paper A. The curves are gentler and therefore indicate that the colour green is less intense (dominant wavelength is 520nm). Unlike paper A, exposure to light in this case led to a less obvious yellowing of the paper due to the loss of the intensity of reflection in the blue area of the spectrum. This effect also makes the paper turn yellow, but to a lesser extent than the original paper in case A. That is why paper B is now seen as being "greener" than paper A, though originally it would have been the other way round.
   
   
3. In both cases, yellowing of the paper has led to an increase in luminosity and dominant wavelength: they become lighter and yellower.

Those data clearly suggest that the two types of paper chosen by the artist had different pigments, which have evolved very differently over time and under the effect of light. The appearance of the work at the time it was completed must have been like the negative of its current state: paper A would have been an intense green and paper B would have been a pale green.

Diffuse Reflectance Data at Each Reading Point

Sample reference number: 67-A

Sample description:
The sample is a piece of dark green paper. The aim of the study is to establish the paper type.

Analysis:
The sample was analysed on the basis of staining and observed under an optical microscope, which allowed us to determine the probable origin of the paper. In this case, the most probable origin of the paper is chemical pulp mainly from resinous plants.

Optical microscope image of this paper sample's fibres, in which the resinous plant fibres can be observed.

Sample reference number: 67-B

Sample description:
The sample is a piece of light green paper. The aim of the study is to determine its origin.

Analysis:
The sample was analysed on the basis of staining and observed under an optical microscope, which allowed us to determine the probable origin of the fibre from which the paper is made. The most probable origin of the paper is chemical pulp with fibres from resinous and leafy plants.

Optical microscope image of this paper sample's fibres, in which the fibres from leafy and resinous plants can be observed

Detail of a leafy plant fibre from this paper sample