Volume 5, Number 1
Mar 1992

Permanent Papers--From Then to When?

Reprinted with permission from the AICCM Bulletin 16 no. 4, p. 25-32, 1990, a special issue entitled "Conference Proceedings: Permanent Paper." AICCM is the Australian Institute for the Conservation of Cultural Material.

The problem of permanence in paper is by no means a new one. Even as early as 400 AD, Chinese officialdom ordered that only paper prepared from rattan and mulberry bark could be used for official documents. In Europe, paper was also regarded as suspect. It was forbidden for official use in Germany by Emperor Frederick II in 1221. Although some of the European distrust of paper may have reflected a suspicion of anything emanating from the Arab world at the time of the Crusades, it was clearly recognized long before the advent of modem pulping and papermaking methods that paper was not always an ideal archival material. Thus, a simplistic approach to the formulation of permanent papers, based on the assumption that "the old ways are always the best ways" may not necessarily give the desired result.

Paper was not originally developed as a medium for printing, writing and graphic arts. Its earliest uses seem to have been for what we would now call packaging, as an alternative to bark cloth. Indeed, there is some reason to suppose that bark cloth, prepared from strands of bark by pounding the wet strands and forming them into layers, was the immediate forerunner of conventional paper. Some time later, it was found that paper could be an acceptable substitute for the silk cloth then used for writing.

At the time of the introduction of papermaking into Europe over the period 1050 AD to 1500 AD, and for many years thereafter, European-made paper was destined for writing and printing end uses. Thus, the choice of fiber would have been the nearest available approximation to that used for printing and writing papers in the well-established Asian papermaking tradition. In the absence of mulberry and rattan bark, this meant that the predominant fibers in European papermaking were derived from the same source as the cellulose-based textiles and cordage of the day, principally flax, hemp and cotton. This was mainly obtained from used textiles and cordage. These fibers are mainly low in lignin and non-cellulosic polysaccharides and high in cellulose, and the permanence of papers prepared from them is usually quite good.

Some time later, there was a transition from the Chinese practice of only using one side of the paper and a preference for translucent papers to the European system of using both sides of the paper and a consequent need for opacity in the sheet. This led to the introduction of opacifying fillers and, given the widespread distribution of chalk deposits which were already being utilized, especially in England, the natural choice for a filler material was calcium carbonate. Thus, the foundation was laid for the eventual conclusion that permanent paper is best produced from lignin-free, high-cellulose fibers with a calcium carbonate filler. However, it must be recognized that, for many hundreds of years, most paper produced in Europe was made according to this "recipe," so we do not really have any alternative formulations to provide effective comparisons.

New Processes

The next group of changes were all brought about by the increased demand for the clean rags which were the main source of papermaking fiber. The shortage of rags had reached a stage where some quite bizarre sources of rags were explored. The use of grave cloths obtained by raiding old cemeteries was discontinued after it was suspected that such rags carried bubonic plague. On another occasion, it is fairly reliably reported that at least one shipment of mummies was imported from Egypt so that the wrappings could be used to augment the regular supply of rags. Serious efforts were soon under way to produce papermaking fibers from wood and straw. Obviously, a good reference library would have been useful at this point as the Chinese had, by then, been making paper from rice straw for at least 800 years. Nevertheless, by the end of the 19th century, the pulping processes commonly used today to separate wood and agricultural residues into their constituent fibers had all been developed and the products had virtually superseded textile fibers in the paper industry. By the early 1800s, pulp was being produced from straw by boiling it in alkali, although at first the pulp was often overbleached and paper made from it deteriorated quite quickly. Groundwood pulping was first developed in Germany in 1844, but does not appear to have been used commercially until 1870. The first chemical wood pulping patent was issued in 1854 for treating wood with sodium hydroxide under pressure, with patents for sulfite pulping following in 1867 and for kraft pulping in 1884. Thus, both mechanical pulps, which contain virtually all of the original lignin in the wood, and chemical pulps, which have had most of the lignin removed, appeared at about the same time and were often considered together under the generic term "wood pulp."

More recent developments in chemical pulping have largely centered around process optimization, although special mention should be made of efforts to provide an environmentally acceptable alternative to the kraft process. Here, soda-oxygen and soda-anthraquinone pulping have been extensively studied and are commercially available.

On the mechanical pulping side, there has been considerable activity since the first introduction of stone groundwood. Since 1945, changes in equipment design have enabled chips to be used as a raw material instead of billets of solid wood, and many varieties of pulp have been produced, each with its own characteristics depending on the combination of temperature, chemical and mechanical treatment and equipment. These pulps are often known collectively as high-yield pulps, and all contain some residual lignin.

Changes in bleaching techniques have also taken place. Prior to the discovery of chlorine, the bleaching process was restricted to exposure of the rags to sunlight followed by washing with potash or soda ash and acid. Between 1790 and 1930, hypochlorites were effectively the only bleaching agents used. Chlorine bleaching was introduced about 1930, with chlorine dioxide and peroxide appearing between 1940 and 1950 and oxygen even later. The advent of these alternative bleaching agents has led to the use of multi-stage bleaching. This, in turn, has made bleached kraft pulp a viable alternative to the more easily bleached sulphite and soda pulps.

And New Problems

We can now identify several critical events which have had a significant impact on the quality of the paper produced thereafter.

In considering paper permanence at any point in time, it must be emphasized that there has always been a considerable variability in quality. Our present need is to identify more specifically those factors which have a deleterious effect on quality, so that such factors can be effectively avoided both in practice and when describing the relevant requirements in standards and specifications.

Prior to 1790, the pulp fibers used in European and North American mills for the manufacture of white paper were all derived from white rags, either cotton or linen, bleached by alternate immersion in alkaline and mildly acid solutions. If the cloth was not subsequently washed free of acid, the fibers could be seriously weakened. Where fibers from new and used rags have been compared, the new rags have invariably been found to give the more permanent paper. After the introduction of hypochlorite-type bleaching in 1790, further problems arose because of fiber weakening. In the most extreme cases, rags could have been bleached twice, once at the cloth stage and again on conversion from rag to pulp. If the bleaching was carried out under less than ideal conditions, severe fiber damage could result. In one case reported, a Bible printed in 1816 and never used was "crumbling literally to dust" only 13 years later. A more recent example of the effect of process conditions is found in a study carried out at the University of Toronto. Cotton linters pulp was bleached with sodium hypochlorite at pH 7 and pH 11. After 23 years, the brightness of the pulp bleached at pH 11 had dropped from 94 to 67, whereas the brightness of the pulp bleached at pH 7 had dropped from 91 to 22. Over the same period, a softwood kraft pulp bleached with a six-stage bleaching sequence only dropped in brightness from 91 to 84. Clearly, any specification which merely stipulates the type of fiber to be used, without ensuring that the fiber has not been damaged at some stage during processing, is inadequate.

In this context, damage may not be immediately obvious. The initial treatment, particularly where it involves a bleaching step, may merely sensitize the fiber so that it is susceptible to subsequent hydrolysis or oxidative degradation. The variability in the present strength of rosin-free rag pulps produced in the earlier decades of the 18th century (presumably the original strengths were at least adequate for the purpose and comparable with each other) indicates significant differences in susceptibility to the factors involved in the aging process. A satisfactory specification must include a means of detecting or predicting the effect of such differences, even in pulps which are traditionally acceptable as constituents of permanent papers.

During the period 1850-1870 several critical changes in the papermaking furnish (the mixture of fibers and other materials which is fed to the paper machine) occurred simultaneously. Both groundwood and chemical wood pulp became readily available to augment or replace rag fibers. Sizing by adding rosin and alum to the paper stock replaced tub sizing, by which the finished sheet was immersed in animal glue or gelatin (and perhaps hardened with alum).

It soon became obvious that under the process conditions then in use, the permanence of groundwood-containing papers such as newsprint was extremely poor. What those early machine conditions were is open to some conjecture. As papermaking wisdom at the time stated that there was no problem which could not be fixed by adding more alum, the whole system was likely to be quite acidic. A further incentive to maintain acidic conditions would have been, and still is, the tendency for groundwood pulp to turn brown in alkali.

From the evidence of the poor permanence of groundwood-containing newsprint, it was concluded that the fibers in permanent papers must be lignin-free. Indeed, some reports maintained that permanent papers should contain no wood fiber, either groundwood or "purified" (chemical wood pulp). However, several studies report that unbleached sulphite pulps were investigated and found not to suffer serious degradation on artificial aging.

Overall, in assessing the influence of lignin on permanence, there are several significant gaps in our knowledge. We know little about the permanence of lignified fibers in an alkaline paper, about the behavioral differences between hardwood and softwood lignin, about the differences in the properties of residual lignin after pulping using different processes, or about the effect of residual lignin after processes like peroxide bleaching which decolorize lignin but do not remove it. Under some circumstances, a specification can err on the safe side by prohibiting the use of lignified fibers of any kind-the penalty being the elimination of some fibers which might be quite satisfactory and less expensive.

However, specifying "lignin free" fibers could lead to the use of weakened over-cooked or over-bleached pulp. "Bleached chemical pulp," on the other hand, would include pulps produced by processes that give fibers with completely unknown aging characteristics. Specifying a particular lignin content leads us further into unknown territory, firstly because of the uncertainty associated with the analysis of lignins modified by pulping and bleaching and secondly because of our total lack of knowledge of the aging characteristics of differently modified lignins. Setting an acceptable level for lignin could also be misleading whenever mixtures of fibers having different lignin contents are used in the papermaking furnish. For example, a paper containing 10% groundwood fibers and 90% satisfactory low-lignin fibers may behave quite differently from a paper in which 90% of the total lignin has been removed from every fiber. The other descriptor often used is that the paper should be free of groundwood. Given the variety of high-yield pulps now in common use, such a statement is meaningless. Unfortunately, there is not sufficient information available to allow us to decide where to draw the line between "chemical" and "mechanical" pulps-or even whether one should distinguish between the two at all.

Apart from cellulose and lignin, there is a third component of wood substance (and of groundwood) which warrants serious consideration. The so-called "hemicellulose" fraction of wood comprises about 25% of the total-the same proportion as the lignin--but the only recognition which seems to be given to the potentially harmful effects of hemicellulose is an occasional reference to the need for a high cellulose content or a requirement that wood fibers should be completely excluded from permanent papers. However, the effect, if any, will probably depend on the composition of the polysaccharide and the nature of any modifications which it has undergone during pulping and especially bleaching. Again, the available information is insufficient for us to write a descriptive specification which will distinguish between those hemicelluloses which are stable and those which are not.

When looking for evidence relating to the influence of lignin and hemicellulose on permanence, it would be extremely useful to know the chemical composition of those Chinese papers which have lasted for many hundreds of years. Analysis of hemp, flax, jute, ramie and bamboo fibers shows that they contain significant quantities of hemicellulose and, in the case of jute and bamboo, quite a lot of lignin also. Although information on early methods of fiber separation is very scarce, it seems unlikely that all of the lignin and hemicellulose would have been removed by the relatively mild treatments available. Admittedly, many samples of old paper have been adjudged lignin-free on the basis of the phloroglucinol test, but this is notoriously unreliable. The most likely situation is that these long-lasting fibers do contain some lignin and hemicellulose, but these have not been adversely modified during processing.

On the other hand, the poor permanence of newsprint containing large quantities of groundwood has created a body of opinion to the effect that the presence of any groundwood, and by extension any high-lignin fiber, is bad. Nevertheless, some of the results obtained by W.J.. Barrow in 1963 in his study of 19th century papers cast some doubt on this piece of traditional wisdom. Barrow examined 50 book papers from each decade, but the decades which are relevant here are the 1880s and 1890s. On testing the papers from these two decades, Barrow found that of the three groundwood-containing papers from the 1880s, one was the weakest, one was in the middle of the range and one was the strongest. The two groundwood-containing papers from the 1890s were ranked in the top four of the decade in terms of strength. To paraphrase Barrow's conclusion, "Wood fibers undeservedly received the greatest share of the blame for the degradation of paper during that period."

At least part of the reason for these contradictory results is that groundwood pulp, chemical wood pulp and alum-rosin sizing were all introduced at about the same time. This transition took place between 1850 and 1870 and led to a situation described by one librarian in 1890 in the following terms: "Centuries hence, some bibliographer will construct an ingenious theory to explain why no books were printed between 1870 and 19-, the date at which we accomplish the destruction of the forests and begin again on cotton." Indeed, there was no shortage of ingenious theories at the time, ranging from the effects of the introduction of gas lighting to the differences in fiber orientation between handmade and machine-made paper.

One trend which is often overlooked in examining paper produced during the latter part of the 19th century is the steady decline in properties such as folding endurance, even in 100% rag papers. This decline can be related to fiber shortening and may reflect changes brought about by the introduction of the continuous Jordan refiner, replacing the batchwise hollander beater. However, it is also possible that, about this time, papermakers realized that by deliberately shortening the fibers, the uniformity of fiber distribution in the sheet was markedly improved and that fiber shortening was the result of a deliberate decision. As the overall decline in properties where fiber length is critical coincided with the introduction of wood pulp, the quite erroneous conclusion that wood pulp was in all respects inferior to rag was reached-a concept which persists even today.

The same logic appears to have been applied to alum-rosin sizing. Properties of the finished sheet declined as animal glue was replaced by alum-rosin, therefore alum-rosin was responsible. Nevertheless, Barrow's results for the 1850s and 1860s do not show any difference in permanence between rosin-containing and rosin-free papers having similar acidities. What alum-rosin sizing did create was the general acceptance of an acidic papermaking system among papermakers. This, in turn, led to the virtual disappearance of calcium carbonate as a filler and the absence of a furnish component with the ability to neutralize unwanted acid from whatever source.

For the Future

At the present time, we are again in a period of rapid change. Just as the Industrial Revolution brought about the introduction of new technology to the pulp and paper industry as part of the overall pattern of development, so too the Environmental Revolution is providing a driving force for the development and introduction of alternative processes and products. In particular, high-yield pulping, recycling and pulping of agricultural residues are seen as ways of reducing the demands on the forest resource. Also, traditional kraft pulping and chlorine-based bleaching may well be gradually replaced by other processes which are perceived to be more environmentally "friendly."

In coping with these changes, we must consider both the stability of the fibers with respect to aging and the initial strength of the papers prepared from them. We must also ensure that our assessment of stability is obtained under conditions comparable with those occurring in actual practice. Thus, fiber stability should be determined using an alkaline system containing an appropriate amount of calcium (or perhaps magnesium) carbonate. There is no doubt that such a system should be mandatory for permanent papers to prevent damage resulting from acid hydrolysis. Unfortunately, most accelerated aging studies have been carried out on unbuffered fiber and there is reason to believe that they are not relevant in practical applications.

Thus, to allow us the greatest freedom possible in choosing components for the permanent papers of the future, the following information must be obtained: (a) the stability under alkaline conditions of any currently available pulps which would in all other respects be adequate for the intended use, (b) the comparable stability of newly introduced pulps and (c) sufficient data on the physical properties of pulps prepared by new processes or from alternative fiber resources to ensure that the introduction of such pulps will not adversely affect the performance of products prepared from them. In this context, stability is often assessed only on the basis of strength loss, but color reversion should also be considered whenever good print contrast or accurate color rendition should be maintained.

The program outlined above represents a long-term plan with a continuing component which will be necessary for as long as new types of pulp are being developed. Eventually, the information generated should provide a guide for the formulation of permanent papers which is more soundly based (and hopefully less restricted) than is the case at present. However, a guide and a specification are not the same thing and we urgently need a specification which will ensure that paper prepared according to the requirements contained therein will always merit the epithet "permanent." Neither of the approaches currently being taken by the specifying authorities is demonstrably completely reliable. The national standards currently in place generally rely on a general description of the type of pulp to be used, but there are so many possible combinations of fiber source and pulping and bleaching conditions that even the quite stringent requirements of these standards do not cover every eventuality. Early drafts of the international (ISO) standard included an accelerated aging test, but the committee was unable to decide on either the duration of the test or the permissible reduction in strength. Thus, we are clearly still unable to predict actual performance from accelerated aging tests.

Clearly, there is more work to be done before we can be completely sure that our "permanent" papers will be really permanent. Meanwhile, we should recognize that the situation has markedly improved in recent years, but further improvement will require time and money for the necessary research.