Introduction

Accelerated ageing tests are widely used in conservation research, either to predict how long a paper would last or to what extent a certain treatment would prolong the period during which an object, or its material, will keep the qualities needed to be used for its intended purpose. In the context of a library and archive the “object” is a book, an archival file or a piece of art, the “material” is mainly paper.

Accelerated ageing is to expose the object in question to a higher level of energy, mostly heat, but also to light or other radiation and aggressive pollutants. It is presumed that the higher energy promotes the processes that together interact to produce symptoms recognised as “ageing”, that is harming a material’s useful qualities. Moreover it is presumed that these processes take place in the same relationship to each other whether at a high or low energy level. It is also presumed that the reaction, (i.e. the ageing), of a material is proportional to the level of energy used to age it, that is the higher the energy level, the more rapid the decay, and this in direct proportion.

It is well known that these last two assumption are not true.

Cellulose Decay

Among the processes that, together combine within a material to cause the deterioration recognised as “ageing”, hydrolysis and oxidation prevail. In a practical assessment, which professional chemists may consider to be an over-simplification, but which might, however, be helpful to conservators, their pattern can be explained as follows.

Hydrolytic decay is quite simple. If there is a surplus of hydrogen ions, i.e. in acidic surroundings, the oxygen bridge between two cellulose monomers are attacked and broken; the hydrogen ion is connected to the oxygen of the bridge thus forming a hydroxyl group. In this way the concentration of hydrogen ions is reduced below the level specific for the pH governing the system, water molecules are split into hydrogen and hydroxyl ions, this latter becomes surplus and immediately connects to the “open end” of the neighbouring monomer. Only H⁺ and OH^- = H₂O is consumed; in terms of chemistry the acid, which is necessary for the process, catalyses it. This process might also happen at the oxygen in the pyranose ring; the result is comparable, i.e. breaking off the chain molecule, reducing the degree of polymerization.

Oxidation of cellulose is a much more complicated and diverse process. Models for it are to be found in the technical literature [1]. Without discussing them in detail, it can be stated that, which oxidized groups are formed, whether one or the other prevails, is highly dependant on the conditions within the system, i.e. the paper, and among these conditions temperature is important.

As a result of some oxidation processes keto- and aldehyde groups are formed. These groups are highly reactive; they are prone to crosslinking, which is the third chemical process of cellulose decay. This also, most probably, depends on the conditions within the system, i.e. the paper, and again temperature is important.

Doubts

The only aim of this excursion into cellulose chemistry is to demonstrate that the assumption on which accelerated ageing is based is highly unlikely, because the chemical processes that occur together in parallel and which can influence each other to cause “ageing”, take place at different temperatures in different ways.

More doubts in the reliability of accelerated ageing may arise form looking at the two “rules of thumb” that are found in the technical literature on conservation research. The first is:
 

The two statements would be in accordance if the increasing rate of rule 1 was 2.54 instead of 2 or if the temperature rate provoking the double speed was 7.4 instead of 10. But this is playing with numbers. Both rules of thumb are obviously wrong.

The third attempt to provoke and support distrust in accelerated ageing tests arises from a closer look at the Arrhenius relationship, which is often quoted in the technical literature as a scientific base for accelerated ageing. There are several written forms of the equation, one of them is: ln(k) = ln(A) – (Ea/R)*1/T, where

A = the „frequency factor“, a constant,
R = the gas constant, pertinent to ideal gases,
Ea = the activation energy, a constant specific for a certain substance,
T = temperature (expressed in degrees Kelvin),
k = amount (concentration) of a specific substance changed per time unit.

The doubts in the usefulness of the Arrhenius relationship for accelerated ageing of paper arise from common sense:

In light of this closer look the question arises whether the application of the Arrhenius equation to predict the useful lifetime of paper is appropriate.

Use of Accelerated Ageing

Accelerated ageing has been the object of intensive research. Older results have become Standards [2]. From the fact that there are four, and that they are hardly ever used as they are given, but always with variations, it becomes evident that there are many and grave doubts in the reliability of ageing tests. Just recently three new and very carefully made research reports [3], [4], [5] have been published, they give sound and carefully considered ideas on how results of accelerated ageing tests may and must be understood. In one of these reports [5] the purpose of accelerated tests is also defined, i.e. – purpose no.1 – to classify papers into stability groups; stable and unstable, and possibly a third group, moderately stable, between these two; and – purpose no. 2 – to gain a certain idea on the long term effect of a conservation treatment. From the viewpoint of practical paper conservation only the second one is relevant.

Similarities and Differences between Accelerated and Natural Ageing

In the new literature it is stated that there are some, or even many, processes that take place during natural as well as during accelerated ageing. In both, e.g. hydrolysis, oxidation and crosslinking take place. This might be seen as a banal observation. More pertinent is that also the final degradation products are qualitatively very much the same: some mono-carbon acids, such as lactic, acetic and formic acid, some dicarbon acids such as succinic and mainly oxalic acid. The proportions of these final degradation products will differ depending on the temperature and the relative humidity of the system, and most probably there will also be distinct differences of the intermediates: hemicellulloses, oxycelluloses, more or less acidic poly- or oligosaccharides like alginic acid. etc. An interesting statement of the new research is that there seems to be quite a similar reaction mechanism in the temperature range of 70 to 90ºC, but that the reaction mechanism dominating at temperatures below 70ºC seems to be quite different.

Systems in Practical Use

As said above, there are several standards for performing accelerated ageing tests, and again as said above, they are rarely used as given. In the technical literature the following parameters are found as promoting the speed of the processes that, occur in parallel and combine or interact to form what we call the “ageing” of a material:

For research in mass deacidification, as it has been done quite intensively during the last 10 years most often 90°C 50% RH for 12 days has been used. The variant of a regularly changing humidity suggested by Käßberger [3] (80°C, 30/60/ 30% RH, 1 h cycle) is not totally new; it has, for example, been used with other figures, i.e. 90°C 35/80/35% RH and changing cycles of 3 hours, for the research on fighting corrosive ink done during the last eight years in the Netherlands Institute for Cultural Heritage in Amsterdam.

The joint Canadian and US American research project [5], which is aiming to define stability classes, suggests a totally new method of accelerated ageing, i.e. to age paper samples in airtight sealed glass tubes with an internal volume of 145 ±5 ml at 90°C. In order to maintain the aspired relative humidity inside the heated tubes the paper inside the tube must be packed quite densely: 4 g in each tube, that means 37 strips 12 x 1.5 cm for fold measurement or 42 square sheets 4 x 4 cm for tear measurement in the case of paper weighing 60g/m², more paper is required for thinner and less for thicker papers. The starting point for this new method of accelerated ageing is the discovery – also this not totally new – that paper sheets inside stacks, within the book block, e.g., suffer from accelerated ageing more than the outside sheets and more than sheets, which hang freely. The reason for this is simple, final degradation products mentioned, such as lactic, acetic, formic, succinic and oxalic acid, are volatile, at least at high temperature. They can evaporate from free hung sheets and from the outside sheets of a stack, but they remain and provoke further decay in the inner block, at least for the relatively short periods that are used for accelerated ageing. Whether this is true also for the much longer periods of natural ageing, or whether it is more likely that these acids diffuse equally within the stack, i.e. a book block and even out of it, might be the topic of further research. Obviously these acids remain in the airtight sealed tubes and provoke further decay. It might be doubted that this would be near to natural ageing. Another problem of this newly suggested method might be to provide appropriate tubes. These must be made of a material which is not only be airtight, but also chemically resistant and thermally stable, etc.

The most Appropriate Method

The first method of accelerated ageing is more than 100 years old. Later many others were developed, most of them during the last 50 years and the last one just recently. In the publication of Henk Porck [4] there is list of publications, which refers to the question of which method is the most apt; there are 80 relevant titles, since 1980, and Henk Porck’s list is not even complete. Two new methods have been suggested just recently, and there is no reason to suppose that they will be the last. Neither the older nor the new tests really imitate natural ageing, and again none of them provoke other reactions than those that, completely all exist and combine to form what we call “ageing”. The differences between the reactions in all these methods of accelerated ageing are negligible in comparison to the difference between these reactions under artificial ageing and what actually happens during natural ageing.

There are two publications – possibly there are more – where two fundamentally different methods of accelerated ageing, i.e. dry heat at a temperature over the boiling point of water and moist heat at distinctly lower temperature, have been done on several same pairs of paper [6], [7]. From the vast amount of data given it can be seen that the course of changing chemical, mechanical and optical properties of paper provoked by the two ageing procedures can be very much the same and very different, that a longer ageing period at low temperature can provoke higher or lower change than a shorter period at a high temperature. Parallel decay prevails, but there are many exceptions, and there is no regularity in them. A look at the DP measurement (Figs. 1 and 2) which represents one single chemical change, i.e. breakings in the cellulose chain, may support this statement. A look at mechanical strength (Fig. 3), which is influenced by more than just one chemical change within the material, shows even more diversity. Undoubtedly it would be possible to explain them all by analysing the components of the several samples in complete detail and follow their specific chemical changes caused by heat and humidity, and it can not be doubted that the relevant discoveries would be of great scientific interest. However, it seems doubtful whether they would be of any practical use for paper conservation – not to mention the expense of the research.

How to Interpret Ageing Tests

For practical use the following statements, derived from the above reported observations and deliberations, may be pertinent. They are not totally new [8]:

Transferring these statements to paper production, in order to define stability classes, could be done by