Pictures accompanying this article: photo 1, photo 2

   Film archives all over the world have for some time been experiencing a phenomenon - the deterioration of films on safety bases, known as the "vinegar syndrome". This phenomenon rises to significant proportions in a number of film archives, principally those located in hot and humid climates, the problem being more serious than that caused by film on a cellulose nitrate base. Someting has to be done to save the contents recorded upon these deteriorating bases, to prevent a total loss of the information they hold, taking into account the high speed of reactions involved in this process. The purpose of this article is to analyze the mechanical behaviour of safety bases resulting from transformations having occurred in their molecular structure during the deterioration process as well as to discuss the selection of polyester film as a stable receiving base when transferring from bases of deteriorated acetates.

The Mechanism of Deterioration in Cellulose Acetates

In order to establish a relationship between its chemical transformations and its mechanical properties, let us rapidly review the deterioration mechanism in cellulose acetates.

The acetate base reacts with moisture forming a separate acid that, when in certain concentrations, degrades the polymeric chain. The degradation mechanism is hydrolated, producing deacetylation, associated with dehydrogenization involving molecular desaturation.

The reaction has an acidic self-catalytic character which once begun continues at a progressively accelerating rate.

Let us consider the effects of the action of the deterioration process upon cellulose acetate bases as a whole. Degradation of the acetate with a subsequent increase in free acidity creates a completely disastrous acidic medium within the container holding the film. Rupture of the polymer leads to the release and diffusion of the plasticizer and solvents, which besides producing drastic geometric deformations (shrinkage, warping), being free will recrystalize on the surface of the film and/or decompose in reaction with humidity, forming acids which attack both the material forming the image and the gelatine present in the adhesive substrate and in the emulsion. Gelatine submitted to very low pH values will degrade and melt.

There is a stage of deterioration of films on an acetate base that we are concerned with determining, before the melting of the emulsion and/or substrate, where depending upon the degree of alteration the mechanical properties of the base have reached, it is still possible to transfer the information on these deteriorated bases to other stable ones.

Selection of the Samples

In order to verify the capacity that films on deteriorated safety bases would have for utilization, a number of sample films, naturally aged, were selected with varying degrees of degradation.

Cellulose Triacetate

The test samples were selected from positive films, 150µ in thickness, from a number of manufacturers, covering a period varying from the 1950's to the 1970's. We selected films in various stages of deterioration, which through systematic observation we identified as characteristic stages in the deterioration process - an odor of vinegar, slight geometrical deformations, marked geometrical deformations, migration of the plasticizer (crystals deposited on the surface of the film), melting of the emulsion and/or adhesive substrate. We stopped at this stage, as from there on the film has no more practical purpose.

Cellulose Nitrate

We introduced into the tests a sample of film on a cellulose nitrate base, a positive from the early 1920's, with the emulsion melted - and in terms of practical utilization just as deteriorated as the most deteriorated triacetate sample - to provide the possibility of comparing the influence the degradation of these plastics has upon their mechanical response, and to demonstrate, for example, that equipment safely operating a deteriorated nitrate could cause permanent damage to a triacetate that apparently has the same level of degradation. The cellulose nitrate, a pyroxylin plastic, is exceedingly unstable, highly inflammable and absolutely unfeasible as file material.

Polyester

The samples of films with a polyester base were taken from various types of material, so as to cover the greatest possible period of storage time, roughly 25 years. Only material stored in rolls of at least 100 feet were used (microfilms, cinema films, x-ray films in rolls).

History of the Samples

All the samples had been submitted during that period of time to similar storage conditions, with temperatures varying from 20°C to 30°C and relative air humidity varying from 70% to 90%. Over lengthy periods of time (20, 30, 40 years) small variations from these parameters are negligible.

Two samples were also selected as a reference standard, that is, material to be used as a comparison of the degradation of the others: a positive film on a cellulose triacetate base processed in 1990, and a positive film on a polyester base processed in 1991. Both samples had been subjected to the same type of storage as the other samples. The reason for choosing a reference standard over three years old is that cellulose triacetate takes approximately this much time to become stabilized, reduce the evaporation of residual solvents and render more consistent mechanical responses possible.

Determination of Residual Acidity

Because of the acidic self-catalytic nature of the degradation of cellulose acetates, the level of residual acidity is a general indicator of the extension of this reaction and is sufficient for our purposes. The residual acid in the samples was determined by leaving a segment corresponding to 1 g of each one for three hours in vessels containing 100 ml of distilled water, whereupon the pH was measured by a potentiometer. Segments of films in various degrees of deterioration were next submitted to mechanical tests, relating the results obtained to the residual acidity that had been ascertained.

The following mechanical properties shall be examined: tensile strength, break elongation, Young's modulus and flexibility.

Tensile strength

Tensile strength, or breaking tensile strenght, or tenacity, is determined by the load applied to the material by area unit at the moment of breakage. Described in ASTM Methods D412, D638 and D882.

Break elongation

Break elongation represents the percentage increase in the length of a sample under traction at the moment of breakage. The test methods applied are the same as those for determining tensile strength.

Young's modulus

Young's modulus is measured as being the ratio between tension and deformation, within the elasticity limit, in which deformation is totally reversible and proportionate to the applied tension.

Recovery

Recovery represents the degree at which material returns to its original dimensions after removing tension. Recovery is expressed as a percentage of the value of the original dimension. ASTM D 412.

Flexibility

Flexibility, or resistance to dynamic flexure expresses the maximum tension, applied alternately as traction and compression, that a material can withstand when submitted to consecutive bending and unbending. It is quantified by the number of cycles the sample withstands under conditions in ASTM Method D 671.

Preparation of the Samples

Two sets of tests were conducted, one to determine the average responses of the part, for which we employed test samples with an area equivalent to 3500 mm2, four samples being tested in each case, providing a final mean value of the results. For the other set of tests we used 20 mm2 test samples, taken from 7 standard locations in each sample, so as to evaluate the variation in the degradation of the mechanical responses within one same sample - one same frame. In this case the tests were repeated three times, the final result being the mean of the test results.

The 7 standard locations (Fig 1) were selected so as to provide data in areas around the perforations - geometrical deformations, shrinkage, loss of strength, flexibility, etc. - where the film is pulled, in the area occupied by the optic sound track, in the center of the film and on mean lines between the center and the perforations.

Analysis of the results

I felt that the tests conducted would have a better reception if they were done on naturally aged samples, or, to evaluate the responses of naturally aged films. It is certain that the rate of deterioration is a function of storage conditions and that the films used as samples had spent their entire lives under deplorable storage conditions, so that the kinetics of their reactions were accelerated as compared to films stored under good conditions. However, all the data were obtained by direct readings, no extrapolations having been made.

In all the tests it is noted that performance along the edges of the film is always worse than in the center. This is due in the first place to the fact that the edges of the roll of film are more exposed to air, humidity and other reagents than the rest of the film, and secondly due also to the stress produced by cutting and perforating the base during manufacturing. The cutting done during preparation of the test samples also produces a certain degree of stress affecting the performance of the part, so that the data produced in the tests are usable only for comparisons between one another.

Testing with Large Samples (3500 mm2)

A set of tests was conducted relating the chemical degradation of cellulose triacetate with alterations in its mechanical properties.

Fig.2 shows that the increase in acidity, a phenomenon related to the degradation of the polymer, significantly reduces the part's tensile strength. Degradation of the polymer is also reflected in the dimensional stability of the material. Fig.3 shows a direct relationship between an increase in acidity and an increase in shrinkage of the part. Degradation involving the loss of plasticizer produces a drastic reduction in elasticity, as can be seen in Fig.4.

The combination of two characteristics of deteriorated films - tensile strength x shrinkage - makes the more serious cases of deterioration less vulnerable than the cases where the deterioration is not so serious, as films that have shrunk excessively require special equipment.

There is a certain tendency to draw a comparison between the shrinkage of a cellulose nitrate base and that of a cellulose triacatate base. The mechanisms that produce the shrinkage of these two bases are essentially the same, but as we will see, the nitrate upon deterioration preserves its mechanical properties far better than does the triacetate under the same conditions. Note in Fig.3 that the shrinkage of the nitrate base is of 4.6%, while that of the most shrunken triacetate is 3.8%, although its tensile strength is 5 times less than that of the nitrate tested. Therefore, extra care should be taken when handling deteriorated safety films, as even with relatively little shrinkage (1 - 2%), the degradation of its mechanical properties is sufficient to cause permanent damage.

Degradation of the base is not uniform. It is more marked at the ends of the rolls and close to the edges, but it is difficult to imagine the extent of this difference in a cross section of a frame, for example. There are cases such as the ratio between maximum and minimum recovery in one same sample reaching a figure of 500% as can be seen in Fig.6, sample 5, or where the mean deforming load is very low (0.02 KgF/mm2) as in Fig.7. In these same figures one may see that the polyester base, sample 1, even after the passage of almost 15 years, maintains excellent homogeneity of response to recovery, showing practically no difference in maximum and minimum values, and even with its strength slightly affected, it requires a load 20 times larger than sample 5 to suffer permanent deformation; even when compared to the reference triacetate, the polyester withstands 1.6 times more load before flowing.

There is a certain apprehension about using polyester which derives from the fact that it has a high breaking tensile strength. This strength is higher when compared to triacetate (approximately 70% greater) but it is not so high when compared to cellulose nitrate (33% greater). Please note that the nitrate sample in question is deteriorated.

Figure 8 shows a comparison between the breakage tensile strength of the various samples tested from one end to the other. We have - by order of tenacity - polyester, cellulose nitrate, the reference cellulose triacetate, slightly deteriorated triacetate (3% shrinkage, pH 3.2) and finally badly deteriorated cellulose triacetate (3.8% shrinkage, pH 2.8). Figure 9 shows the percentage variation between maximum and minimum values of each sample. Once again, the polyester shows more homogeneous responses, the nitrate taking fourth place with a 31% variation. In this set the sample with the least homogeneous tenacity is number 5, with a variation of 160% between maximum and minimum within the area of one same frame! This difference makes it impossible to copy this material by contact, as the required pressure would be far greater than the permanent deformation load, and would produce generalized flowing in the material.

One factor that contributes for polyester to be preferred as a base for cinema films over other less chemically and physically stable materials is based upon the extremely elastic characteristics of its break elongation. As can be seen in Fig.10, the deteriorated cellulose nitrate elongates almost the same as the reference triacetates (roughly 50% of the initial length). It can also be noted that deterioration causes this characteristic to fall to 10 - 5% in the case of cellulose triacetate, whereas the polyester elongates to around 160%!

Accelerated Tests with Samples

Samples of film on a polyester base showed almost no change in mechanical behaviour amongst one another. We submitted two samples of film on polyester and cellulose triacetate from the same rolls as the reference samples to an extremely acid condition (submersion in concentrated glacial acetic acid); the cellulose triacetate for one hour and the polyester for 10 hours. Hydrolysis of the triacetate was so strong that the sample suffered a lateral shrinkage of almost 10%, while the polyester showed no dimensional change after 10 hours of immersion.

20 mm2 test samples were prepared from these samples and submitted to the previously described tests. It may be seen in Figures 10, 11 and 12 that even after being submitted to hydrolysis in an extremely acid medium for a period 10 times greater than the hydrolyzed triacetate, the polyester maintains mechanical response characteristics better than those of the reference cellulose triacetate, whereas the triacetate falls to the levels of the most deteriorated sample in the previous test series.

Conclusions

Our conclusions are that during the deterioration of cellulose acetate films, drastic alterations are developed in the responses to mechanical stimuli which in the end render them unfit for use due to geometric alterations, loss of elasticity and a drop in tensile strength, brought about by chemical degradation.

The differences in mechanical behaviour between deteriorated cellulose nitrate and cellulose acetate bases are considerable, principally in terms of elasticity and tensile strength. Equipment, even when specifically designed for the restoration of films, and which may safely draw shrunken cellulose nitrate film, may cause permanent damage to a cellulose acetate film.

Having established the intrinsic instability of a cellulose triacetate base, and having demonstrated the levels of degradation it can reach during the process, it appears that duplicates and preservation prints made from originals on acetates should be made on more stable bases.

Throughout its history as a base for cinema films, polyester has had deficiencies which have been solved, such as problems related to residual electrostatic charges caused by its low electrical conductivity, those related to splicing due to its insolubility, etc.

There is also the question of its high tensile strength which sometimes may cause damage to equipment, but triacetate also damages equipment. Among the bases available for cinema, film polyester is undoubtedly the most stable - both chemically and physically. I do not think it would be practical for the majority of film archives to even consider passive preservation of more film on cellulose triacetate. Storage conditions for polyester are more practical, and it would be far more rational to reinforce a few shafts than to provide the dramatic climate requirements to control the cellulose triacetate reactions at the levels required for archival purposes.

Joâo Sócrates de Oliveira

Dans son éloge des qualités du polyester, l'auteur signale que les défauts observés tout au long de l'histoire du film en polyester (charges électrostatiques résiduelles, insolubilité rendant difficile le collage, etc.) ont été corrigés et que d'autres inconvénients (la résistance à la tension qui endommagerait certains appareils) sont largement compensés par ses qualités (stabilité chimique et physique) qui font de ce matériel un support adéquat pour la plupart des archives.