Mass Deacidification of Paper and Books

Mass Deacidification of Paper and Books

II: Deacidification in the Liquid Phase Using Aminosilanes

by E. ROUSSET, S. IPERT, & H. CHERADAME

introduction

In the first paper of this series the feasibility of a gas phase deacidification process using nitrogen containing organic molecules was discussed1. It was concluded that such a process was not easy to develop despite its obvious apparent advantages. Not only must the molecule, which is in charge of neutralizing the acidity, contain the necessary basic group, but it must also be able to react with and to bind covalently the molecule to the cellulosic material so that pollution problems linked with the necessary alkaline reserve, and environmental contamination can be avoided. Molecules meeting these requirements and which are volatile enough to be used in a gas phase are not common. This is due to the fact that cellulose does not easily react and it is difficult to bind the reactant covalently. It was therefore decided to investigate processes by which the reactive molecule would be dissolved and the material impregnated with a solution. Of course, it is necessary to dry the material after treatment to remove the solvent. Water cannot be used, not only because any aqueous treatment of whole books is highly problematic, but also because water can destroy the molecule in charge. Taking into account the chemical nature of the cellulosic materials, it was assumed that a suitable solvent could be found from the alcohol family. Another fact that has to be taken into account was the recently regretted lack of commercially available non-aqueous deacidification processes using solutions free of chlorofluorocarbons and methanol2. In view of the preliminary investigations, it seemed promising to try a reactant from the aminosilane family1, some of which being soluble in ethanol.

Deacidification processes using amines are not common. The only real example is the Japanese process based on a combination of ethylene oxide and ammonia, in situ producing ethanol amine3. The main reason for the lack of this kind of process is linked with the rather high volatility of these types of compounds. In connection with the use of organic compounds based on nitrogen, the use of an alkaline derivative of diazomethane to esterify the oxidized parts of the cellulosic

substrate is currently being researched4. The neutralization of acidic oxidation products of cellulose was an interesting idea. However, the regeneration of acidic groups from the esters, with long term hydrolysis, was considered to be problem.

Our aim was to remove reactants based on alkali cations for reasons mentioned in our previous paper1. Alkaline earth cations, such as magnesium, are too acidic and can cause problems with long term ageing5' 6. It was hypothesized that amine based functions could be of some interest if they remained firmly bound to the network. Since this result could not be easily obtained by a chemical reaction on the cellulose, it was decided to check the reaction using a self cross-linking amine, such as the aminosilanes studied in this paper. One possible drawback of using amino-based compounds could have been the formation of small amounts of nitrous compounds in the long term. Indeed, the carcinogenicity of some nitrosoamines has been established7' 8. However, the fact that the nitrosoamines cannot be produced by primary amines and that these side-products would be covalently bound to the fibre network and consequently not accessible to the environment, is probably an adequate response to the contamination problem.

This paper is devoted to the evaluation of the effect of amino alkyl alkoxy silanes introduced into naturally aged acidic papers.

experiment

Acidity measurements

The acid content of papers was determined as previously described1. Two measurement techniques were used. Most frequently surface pH measurement was used: a drop of distilled water was deposited onto the paper surface and the contact with a flat glass electrode allowed to directly read the pH from a pH-meter once the indicator had stabilized, i.e. after ca. 30 sec. This technique rapidly provided an indication of the acidity of the samples, but its exact meaning must be questioned. According to Fujiwara et al.9, the extraction of paper in an aqueous medium can lead to the hydrolysis of some compounds from, for example, the sizing, and present only in the surface region of the paper and would not therefore be representative of the material as a whole . Thus, a more accurate measurement of acidity was determined using aqueous extraction according to the standard procedures10.

Alkaline reserve was determined according to a similar procedure11. It was sometimes found that, when the alkaline reserve content was low after treatment (and/or ageing), the cellulosic material had both a small amount of acidity and a small alkaline reserve simultaneously. This phenomenon will be discussed later. It

Table 1: Qualities of the papers (average of at least 8 measurements on the same sample).

led to the conclusion that the alkaline reserve must only be considered as being well established when it amounts to at least 20meq / lOOg, which corresponded to 1% CaCO3.

Origin of the acidic papers

It was decided to study the new deacidification process on naturally aged books consisting of acidic paper and originating from the beginning of the XXth century. They were obtained from the antiquarian book market. This procedure had the advantage of dealing with a real situation within a library, allowing to look also at the effect of the treatment on the bindings. The surface pH of the paper was in the range of 3.5-4.5. Some characteristics are described in Table 1. Accelerated ageing after the deacidification treatment was done by heating in an oven connected to open air for a given time, unless otherwise stated.

It should be noted that ageing at 80°C for 20-26 days did not significantly modify either surface pH or acid content of any of the books. These two characteristics do not directly correlate. Paper F and I, e.g., which have the smallest acidity content also have the lowest surface pH, which illustrates the limitations of measuring only the surface pH to determine acidity.

To be certain, it was verified that the deacidification treatment did not influence the results of the determination of the fibrous composition. Consequently, this analysis could be carried out before or after treatment.

General procedure for treatment

Preliminary experiments showed that most papers could absorb an average of 80% (w/w) of ethanol depending on the experimental conditions. This allowed us to calculate the concentration of the reactant in order to determine its theoretical content in the paper after treatment and drying. Paper impregnation can depend upon the contact time. Unless otherwise stated, a contact duration with the treating solution of 10 min was chosen.

The general procedure for paper impregnation was as follows. The various aminosilanes were dissolved in anhydrous ethanol at the required concentration, and the treatment of the papers by immersion was carried out in a glove box under nitrogen, unless otherwise stated. The samples (8 pages) were then placed in a glass dessicator connected to a water trump for solvent evaporation. PH, weight gain and some other characteristics of the samples to be reported in a follow-up paper were determined, which were, in the meantime, kept in a conditioned environment (24°C, 45% RH). For the treatment of complete books, the procedure was similar, but the equipment was adapted to the size of the document.

Silanes used

Five different silanes were used:

• 3-aminopropyltrimethoxysilane:

NH2-(-CH2-)3-Si(O-CH3)3 (ATMS) *;

• 3-amino-2,2-dimethyl propyl tri-methoxy silane:

NH2-CH2-C(CH3)2-CH2-Si(O-CH3)3 (AMTMS) **;

• 3-(N,N-dimethyl amino)propyl tri-methoxy silane:

(CH3)2N(-CH2-)3-Si(O-CH3)3 (DMATMS) **;

• 3-amino propyl tri-ethoxy silane:

NH2-(-CH2-)3-Si(O-CH2-CH3)3 (ATES) ***;

• 3-amino propyl methyl di-ethoxy silane

NH2-(-CH2-)3-Si( O-CH2-CH3)2-CH3 (AMDES) ***.

These compounds were used as received from the producer, without further purification. They were dissolved under nitrogen in a glove box in absolute ethanol, or in ethanol at 95% when specified.

*Aldrich

**Witco Corp.

***Gelest ABCR

Table 2: Effect of treatment with pure ethanol on paper D.

*Measured using a spectrophotometer (Elrepho 2000) which determines reflectance.

**Measured on samples of 50 mm length and 15 mm width with an Instrom instrument working at a rate of 10 mm/min, according to the usual standards.

results and discussion

Since it was decided that the solvent should be a member of the alcohol family, ethanol was chosen for the sake of simplicity and non-toxicity. The effect of ethanol on acidic papers had been determined previously.

Paper D was chosen for this test. The results are shown on Table 2. As usual pH was lower at the edges than in the centre of a sheet, showing the influence of the environment on paper degradation.

It can be seen that extraction with ethanol was acted as a kind of washing. The ethanol became pale yellow, while the treated paper exhibited a slightly lower yellow index and a slightly higher surface pH. Breaking length was also slightly increased, which can be assigned to a small reorganization of the fibrils. Obviously, washing in pure alcohol resulted in the extraction of some yellow acidic derivatives, but this process was by no means a sufficient deacidification treatment. It will be seen below that it did not markedly influence the characteristics of the treatment by aminosilanes.

Effect of 3-aminopropyltrimethoxysilane (ATMS) on paper acidity

Firstly, the effect of a basic molecule in the series of aminoalkoxysilanes was investigated. The reason for this choice was that, upon hydrolysis, the tri-alkoxysilane function was transformed into tri-hydroxysilanol which must have given a network due to self-condensation. It was expected that not only would there be a dramatic decrease of volatility after hydrolysis, but also a strong fixation on the fibre surface due to the formation of the network, according to the following scheme:

Table 3: Effect of a preliminary alcoholic extraction on the neutralization by 3-amino propyl tri-methoxy silane (ATMS) as shown by the surface pH .

Table 4: Variation of silane retention and surface pH of paper E with various concentrations of the treating solution (ATMS in absolute ethanol).

As a result, the aminosilanes were deposited on the fibre surface, contact with moisture after contact with air or with undried paper should have induced the formation of the silanol groups which would undergo further self-condensation.

The results of this first series of experiments are given in Table 3. It can be seen that, even with a small concentration of aminosilane in pure ethanol, the surface pH increased, and that with a concentration of 2% (w/w), the pH apparently reached the minimum value required for effective neutralization. There was no significant difference between the paper previously treated with pure ethanol (sample no.2) and the non-extracted paper (sample no.l), the small difference in pH between the samples being within the accuracy range (see below).

From the comparison of the results given by the two runs (Table 3) it can be seen that treatment using solutions of low concentrations did not seem to be seriously modified by a preliminary extraction with ethanol. From this fact it can be deduced that most of the acidity was linked to the fibre surface or trapped within the bulk of the cellulosic material. Modifications in the acidity of the papers may be attributed to having contact with aminosilane. This was confirmed by the variation in the surface pH as the concentration of the aminosilane increased. Relevant data are given in Table 4, together with data on the parallel silane uptake. They may help to find the optimum concentration for a satis-

Table 5: Acidity content of a paper submitted to accelerated ageing by irradiation (Xenotest: 35°C, 25% RH, UV irradiation) and by heat (80°C). The tests were done with paper E.

factory deacidification and for providing a sufficient alkaline reserve. Table 4 shows that the silane uptake was lower than the theoretical uptake, which could be deduced from the weight uptake (= 80%) of the treatment solution and its silane content. This lower retention can probably be explained in the same way as described in the recent study carried out by Saint-Cyr et al. where, absorption of cationically charged molecules is governed by an ion-exchange process accompanied by a specific absorption on the fibre surface12. Thus, it is not surprising that the silane uptake is different from the uptake which can be calculated from the solution absorption. This point will certainly deserve special study in the future. However, a preliminary answer to the question of identifying the optimum concentration may be addressed by measuring the alkaline reserve and is set out below.

Effect of various aminosilanes on paper acidity

Three different aminosilanes were used: ATMS, AMTMS DMATMS (see p. 107) They have different chemical structures: ATMS involves a primary amine on a linear propyl group. It is the reference compound because it is industrially produced in large quantities. AMTMS also contains a primary amine group, but on a p-disubstituted propyl group which cannot consequently give β- elimination. DMATMS contains a tertiary amine group, the chemical reactivity of which is different from the primary one.

A comparison was made between ageing in a Xenotest machine and in an oven. The resulting acidity is shown in Table 5. It can be seen that ageing in the Xenotest resulted in greatly increased paper acidity, which supported the well-known detrimental effect of UV irradiation on paper quality. The heat ageing chosen for our experiment did not bring about a significant acceleration of cellulose degradation as far as it was expressed by increased acidity. The question as to which is nearer to simulating the changes taking place in paper during long term natural ageing at room temperature deserves special study.

Table 6: Variation of surface pH upon heating at 80°C after treatment with two different amino-silane solutions (3 w/w % in pure alcohol). The paper was taken from book E.

Table 7: Variation of the surface pH upon heating at 80°C for 14 days after treatment with various aminosilanes (9 w/w % in pure alcohol). The paper was taken from book E.