![[CoOL]](/icons/sm_cool.gif){kind=icon}
This report is based on the status of process development at the time (August 1993) the test books were subjected to treatment by the Bookkeeper Process. At that time, some issues were identified that were thought to be of interest to the library and technical communities. These issues were pursued by requesting that a number of tests be performed on treated test books. The main driving force behind these requests was anticipating potential users' questions and bench-marking efficacy. Finally, the reader should note that PTI is actively engaged in continually assessing and improving the Bookkeeper process. Thus, this section should be viewed as a review of the process' principles and a "snapshot" of its capabilities at the time the test was conducted on August 30, 1993.
This review begins by providing a technical overview of the process. Thereafter, the processing aspects that generated some of the analytical requests are noted to provide the reader with a broader context for reaching conclusions from the results presented elsewhere in this report. This review concludes with comments on process capacity and control and outlines the directions PTI was pursuing at the time of evaluation. The appendices contain a statement written by PTI that describes recent process modifications in more detail.
In this overview, what a typical hard cover book experiences in the Bookkeeper process is described. Overall, the aim of the process is to impregnate the text's pages with alkaline, submicron MgO particles. The MgO particles are dispersed in a perfluorocarbon carrier fluid by using a surfactant. The carrier now in use replaces the CFC compound originally used; the new carrier was adopted to minimize environmental impact. The surfactant is used to prevent the aggregation of particles. If aggregation occurred, particle penetration into the paper's fibrous structure would be reduced. The surfactant contains a perfluorocarbon-functionality which endows it with solubility in the carrier fluid.
The texts to be processed in the batch are first attached to a horizontal, V-shaped platform. Attachment is accomplished by clamping the binding's corners to the platform surface. Thereafter, the platform bearing its load of texts is immersed into the MgO suspension. Immediately after the load is immersed, the air from between the pages and binding exits. The high specific gravity of the carrier fluid (1.7) fosters flotation resulting in the opening of the text's pages. At this point, a text appears to be half open with the fanned out pages supported by the dense carrier fluid. Often, dirt and debris from the books floats to the top of the bath thereby providing some cleaning.
Apart from the materials used, another key component of the Bookkeeper process is the agitation a text experiences following immersion. MgO particle penetration is aided by establishing convective mass transfer. This meansthat the carrier fluid flows over the opened pages. Such motion is intended to increase the rate and extent of contact between the pages and MgO suspension. Carrier fluid motion is induced by moving the platform back and forth in the horizontal plane; motion occurs parallel to the text's spine. This motion establishes mild fluid circulation in the immersion bath. Adjusting the range of platform motion is possible prior to operation. This adjustment regulates the period and phasing of the circulation. The motion combined with the high specific gravity of the carrier fluid causes the pages to fan as the platform moves. The motion of branched kelp in a gentle tidal flow is reminiscent of what is observed. As the flow passes by and recedes, the kelp's branches move laterally back and forth.
Following impregnation for ca. 12-15 minutes in a bath containing 2.5 g MgO/l, the platform is raised above the immersion bath and excess carrier is allowed to drain. Thereafter, the platform and attached books are transported to an evaporator chamber. There, the carrier fluid is evaporated at approximately room temperature conditions. The evaporated carrier is collected and condensed so that it can be recycled. The normal boiling point of the carrier is 80� C; hence, the evaporation process for a ca. 200 page text is complete in less than ca. 16 h based on gravimetric criterion. Carrier recovery observed in the August 1993 process is less than 100 percent.
Issues Relevant to Efficacy Penetration & Uniformity of MgO Distribution
Apart from the type of paper, the operation of the equipment (e.g. duration of immersion), and size of the text, the shadowing effect of neighboring texts could conceivably affect the treatment a book experiences in a batch. If these effects occur, then the MgO particles could prove to be not uniformly distributed on a page or throughout the text.
PTI has considered the MgO distribution issue. At the time of Library-sponsored study, PTI staff utilized alkaline pH indicators to determine if treatment confers alkalinity to all areas of a given page as well as to all pages in a text. Supplemental information in the form of scanning electron micrographs was also provided to the Technical Evaluation Team. The micrographs indicated that particles are adsorbed on cellulose fibers. Elemental analysis via energy dispersive x-ray analysis is also claimed by PTI to confirm that the particles observed in micrographs are indeed MgO. Overall, the operational parameters (e.g. immersion time, platform motion range) used in this study were based on PTI's prior videotaped motion studies and analyses (e.g. pH indicator studies) of their own test books. The results of fold endurance and other tests which bench-mark efficacy are discussed by Paul Whitmore. Process Control & Quality Assurance The treatment is a batch process. Thus, as time elapses, MgO concentration in the bath may potentially change due to uptake by texts or carrier fluid evaporation. The former would reduce MgO concentration while the latter effect would increase concentration. At the time of the test, once the range of platform motion was set, the main method of controlling the process entailed periodically withdrawing a bath sample for analysis. The control strategy involved maintaining the MgO concentration at the level anticipated to yield the desired alkaline reserve for the treatment times used.
Assessing the extent of treatment for actual circulating books is not straightforward. Using pH indicators is obviously undesirable due to the staining; hence, selected surface pH electrode measurements would have to be performed. Alternately, representative samples of low value books could be inserted in a batch and subjected to destructive testing.
Relying on historical data and limited analyses is an omnipresent challenge in the majority of deacidification processes.
Process Capacity and Scale-Up Potential
At the time the tests were conducted, PTI had one immersion bath and one drying chamber. Both units can be regarded to be pilot or semi-works scale. The bath was capable of processing 3-5 books per cycle. The drying chamber was multi-tiered and could handle the product from numerous batches.
We observed that the process has two labor-requiring steps that contribute significantly to cycle time. First, the attachment of the books to the platform consumed 10-30 minutes. The loading of conventional sized hard cover books with reasonably intact bindings was fairly straight forward. Books with weakened bindings required special handling and the use of additional supports. The second labor-intensive step entailed transferring the books from the bath to the drying chamber (ca. 15 minutes). For one treatment bath of the scale demonstrated, we estimate that 125-175 books can be treated per 24 hour day. Adding more treatment baths and increasing bath and drier capacity would raise throughput and transform the operation to semicontinuous. If the scale-up of the treatment tanks was considerable, then the relationship between range of platform motion, inter-book distance, and treatment efficacy would have to be assessed to ensure that the relationships established from pilot studies hold. PTI has acquired some experience with this optimization problem. Drying at a larger scale would require a shift in how treated product is inventoried as well as an increase in carrier recovery capacity (e.g. chilled condenser) beyond that currently available.
One key component of the PTI process is the surfactant obtained from an off-shore, Italian vendor. Because this source is presently unique, the process may be vulnerable to vendor stability or whims unless substitutes or other sources can be identified. Alternately, an inventory strategy must be developed.
To address this concern, PTI has obtained a letter of commitment from the vendor that states that the likelihood is very low that the product line will be discontinued in the foreseeable future. Additionally, PTI has been in communication with 3M Company on manufacturing a perfluorocarbon surfactant. Such a development would represent increased vertical integration of product lines for 3M because they also vend the carrier used in the Bookkeeper process.
At the time of testing, PTI was considering different means for increasing treatment capacity. Additionally, implementing additional process control strategies was under discussion. One potential control strategy entails installing an optical sensor in the bath to monitor continuously the MgO concentration. Conclusions
From the processing standpoint, the Bookkeeper process has the advantage of simplicity. Essentially, two components are required: a controlled solid-liquid contacting system and a volatile liquid recovery/recycle system. Many aspects are amenable to optimization which, in turn, would lower costs and increase throughput. Increasing the utilization of bath volume and the recovery of carrier, for example, would conceivably reduce the cost per batch.
A statement provided by PTI on current work can be found in Appendix E. Overall, the materials and requirement of fluid motion relative to thetext will be fixed characteristics. The statement, however, will allow the reader to gauge how capacity and other features may change over the next year.
Many of us think that science can resolve almost any problem if given enough time and funding. Our expectations are high, and we demand perfection. Several deacidification processes have been developed over the past decade. Over and over again, librarians were assured that the newest process was the ideal solution they had been waiting for. Further testing and research inevitably revealed otherwise. All processes have had drawbacks; some were totally unacceptable; others showed potential. Clearly, however, compromises will have to be made. It is important for those involved to understand treatment chemistry, to educate themselves about collection conditions, and to make informed decisions. This report, written by a paper chemist, contributes to the continuing discussion and analysis of Bookkeeper. Using the data generated through extensive testing in his own lab, by the Library, and by IPST, he has provided insight into the potential of this process to meet the Library's specifications and to serve the needs of the library and archival communities.