ANALYST, MAY 1992, VOL. 117 849 Determination of Neutral Sizing Agents in Paper by Pyrolysis-Gas Chromatography Tatsuya Yano,* Hajime Ohtani and Shin Tsuget Department of Applied Chemistry, Faculty of Engineering, Nagoya University, Furo-cho, Chikusa, Nagoya, 464-0 7, Japan Takao Obokata DIC-Hercules Chemicals lnc., lchihara 290, Japan Neutral sizing agents, viz., alkylketene dimers (AKDs) and alkenylsuccinic anhydride (ASA), in paper were determined by pyrolysis-gas chromatography, including those that reacted chemically with the paper. The peaks of intact AKDs and related ketones in the pyrograms were used as the key peaksfor the determination of AKDs in paper samples containing between 0.025 and 1 .O% of AKDs. The results obtained suggest that almost 75% of the AKDs added are retained in all the paper samples.Further, the relationship between the AKD content and the degree of sizing is interpreted in terms of the possible paper sizing mechanisms. The ASA content in paper was also determined in essentially the same way as for the AKDs, and is discussed in relation to the degree of sizing. Keywords: Pyrolysis-gas chromatography; neutral sizing agent; paper; alkylketene dimer; alken ylsuccinic anhydride As paper is composed of hydrophilic cellulose fibres, it tends to absorb aqueous liquids by capillary action of the inter- and intra-fibre voids. Although this property is useful for filter and blotting paper, printing and writing paper has to be resistant to the blotting of ink to some extent. Therefore, sizing agents are often added to the pulp slurry or applied to the surface of the paper in order to improve the printing and writing qualities by developing resistance to penetration by aqueous liquids.Rosin-alum is one of the most popular and traditional sizing agents. However, it is difficult to preserve paper sized with rosin-alum for long periods of time because of the acidic nature of alum (aluminium sulfate). On the other hand, sizing agents such as alkylketene dimers (AKDs) and alkenylsuccinic anhydride (ASA) react covalently with the hydroxy groups of cellulose to form ester linkages under neutral or alkaline conditions.1.2 Therefore, small amounts of AKDs and ASA can provide strong sizing effects for prolonged periods without destroying the paper matrix. These sizing agents are generally added to the pulp slurry and are partly lost with wasted white water.Moreover, the AKDs retained in the paper are known to have at least three different forms: I ( a ) intact AKDs physically adsorbed with cellulose R-CH=C-CH-R’ ( 6 ) ketones formed by hydrolysis R-CH~-CC=O R-CH=C -CH-R’ I I I + H20 heat R’-CH-C=O 0-c=o I * Present address: DIC-Hercules Chemicals Inc.. lchihara 290, I- To whom correspondence should be addressed. Japan. (c) (3-keto esters formed by reaction with the hydroxy groups of cellulose R-CH=C-CH-R‘ I I 0-c=o By analogy with (a) intact ASA possible forms: heat OH Cellulose + I ____) R’ 0 I II R-CH2-C-CH -C-0-Cellulose II 0 AKDs, ASA in paper could have three 0 II R-HC-C\ I ’0 H2C-C’ C16H33 II I 0 (R = -CH=C-C16H33, etc.) (b) di-acids formed by hydrolysis 0 0 II II R-HC-C\ R-HC-C-OH H2C-C H 2C-C -0 H I ,o + H20 - I II II 0 0 (c) The ester acids formed by reaction with the hydroxy groups of cellulose 0 0 II II I R-HC-C -0-Cellulose - R-HC-C, OH I , o + t H2C-C Cellulose H 2C- C-0 H I1 II 0 0 By examining ASA in paper with infrared spectroscopy, McCarthy and Stratton2 reported that the esterification of ASA with the hydroxy groups of cellulose and the resulting sizing effect were strongly influenced by the drying conditions.Therefore, the discriminative determination of each form of AKD or ASA would be highly desirable in the field of paper manufacture. In order to study the mechanism of neutral850 ANALYST, MAY 1992, VOL. 117 sizing, Roberts and Garner1 determined the content of AKDs in papers by using 14C-labelled reagents.The radioactivity of the paper samples before and after extraction with chloroform was measured in order to distinguish between the reagents that had reacted with cellulose and those that had not. Pan et al.3 reported that ultrasonic attenuation by paper in water corresponded to the amount of AKDs in a paper sheet. Recently, Dart and McCalley4 determined the content of AKDs in paper by capillary gas chromatography-mass spec- trometry (GC-MS) based on hydrolytic extraction followed by quantification of the resulting long-chain ketones. However, the recovery of the AKDs was not quantitative because they reacted with the hydroxy groups of cellulose still present in the paper even after the extraction. In recent work, pyrolysis-gas chromatography (GC) was successfully applied to the determination of small amounts of a polyamide-epichlorohydrin wet-strength resin added to paper.5 In the present work, AKDs and ASA in paper, including those that cannot be extracted with a solvent, were determined by using essentially the same technique.The results obtained are interpreted in terms of the possible mechanisms of sizing. Experimental Materials The AKD emulsion (Aqapel 12), which contains a solid fraction (20%) including AKDs (86%), produced by DIC- Hercules Chemicals, was added to the pulp slurry at pH 8.0 in order to prepare the paper samples. The samples were then dried with a drum dryer at 100 "C for 50 s. The eight paper samples containing various amounts of AKDs are listed in Table 1 together with their Stoeckigt degrees of sizing6 All the samples contained 0.5% of polyamide-epichlorohydrin resin. The ASA (hexadecenylsuccinic anhydride containing 16% octadecenylsuccinic anhydride) produced by the Dixie Chem- ical Co.was also added to the pulp slurry at pH 7.5 to prepare the other two paper samples containing 0.12% of ASA. These samples also contained 20% of CaC03, 0.75% of cationic starch, 0.5% of A12(S04)3-14H20 and 0.02% of poly- acrylamide. In addition, standard paper samples containing known amounts of AKDs or ASA were also prepared for calibration. A prescribed amount of the AKD emulsion was added to two sheets of filter-paper (Advantec Toyo No. 6,90 mm diameter) and then dried at 105 "C. Seven types of standard paper samples with AKD concentrations ranging from 0.004 to 2.3% were prepared.The standard paper samples containing ASA were also prepared in essentially the same manner. All the paper samples were cryo-milled into a fine powder by a freeze/mill (Spex 6700) prior to pyrolysis-GC measure- ments in order to homogenize the samples. Conditions for Pyrolysis-GC The high-resolution pyrolysis-GC system utilized in this work was essentially the same as that described previously.7 A Table 1 Paper samples containing AKDs Sample No. 1 2 3 4 5 6 7 8 AKD dosage (% ) 0 0.025 0.05 0.075 0.1 0.2 0.5 1 .0 Stoeckigt degree of sizing/s 0.1 4.6 21.9 32.0 44.7 91.8 125.0 162.0 vertical microfurnace-type pyrolyser (Yanagimoto GP-1018) was directly attached to a gas chromatograph (Shimadzu GC-9A) equipped with a flame-ionization detector.About 0.5 mg of the milled paper sample, 0.1 mg of the AKD sample or 0.02 mg of the ASA sample was placed in a platinum sample cup and then pyrolysed under a flow of nitrogen carrier gas. The pyrolysis temperature was set empirically at 500 "C, which was sufficiently high to attain almost complete thermal degradation of the main cellulose matrix of the paper and to achieve thermal desorption of the various reagents added to the paper sample. A fused silica capillary column (30 m x 0.25 mm i.d.) coated with dimethylsiloxane or 5% phenylmethyl- siloxane (0.25 pm thick), immobilized by chemical cross- linking, was used. The 50 ml min-' carrier gas flow rate at the pyrolyser was reduced to 1.0 ml min-1 at the capillary column by means of a splitter.The column temperature was initially set at 50 "C and then programmed to increase to 300 "C at a rate of 5 "C min-1. Identification of the peaks on the pyrograms was mainly carried out using a GC-MS system (Shimadzu QP-1000) with an electron impact ionization source to which the pyrolyser was also directly attached. Results and Discussion Determination of Retained AKDs The pyrogram of the AKD sample at 500 "C is shown in Fig. l(a). The peaks in the pyrogram are summarized in Table 2. The main products are intact AKDs and their hydrolysis products (ketones), which are observed in the latter part of the pyrogram. As the AKD sample utilized consists of a mixture of various ketene dimers mainly with C12- and C14-, C14- and C14-, C14- and C16-, and C16, and C16-alkyl chains, the main peaks in the latter part of the pyrogram are assigned to the corresponding four AKDs and the four associated ketones.0 ~ ~~~ 6 : : : : : . . . . . i ; ; : : . . . . . . . . . . . . . . . . . . . . . . . _ . . . . . . . . . . . . . . . . . . . Levoglucosan \ J I I I I 20 40 60 80 Retention time/min Fig. 1 Pyrograms of a paper sample containing AKDs and pure AKDs at 500 "C. ( a ) Pure AKDs: ( b ) paper with 1.0% of AKDs added; and ( c ) control paper. Peak numbers correspond to those in Table 2ANALYST, MAY 1992, VOL. 117 Table 2 Assignment of the main peaks derived from AKDs Structure of alkyl chains, R,R’* Total number of Peak No. Compound carbon atoms Combination Ketone AKD Ketone AKD Ketone AKD Ketone AKD 26 26 28 28 30 30 32 32 0 R-CH=C -CHAR’ II I I * Kctone, R-CH~CCHZ-R’ AKD.0 - C C O 1 I 0 0.5 1 .o AKDs in feed (%) Fig. 2 Relationship between feed of AKDs into the pulp slurry and AKDs retained in a paper sample, estimated by pyrolysis-GC. Solid line: observed relationship. Broken line: hypothetical relationship for complete retention (100%) About 70% of the weighed AKD sample was recovered as the intact AKDs and the associated ketones in the pyrogram under the pyrolysis conditions used, whereas the remaining 30% was mostly degraded to a series of hydrocarbons (up to C,,), which were observed in the early part of the pyrogram, each of which mainly consisted of a doublet corresponding to an alk-1-ene and an n-alkane formed from the alkyl groups in the AKDs. The pyrogram of the paper prepared by feeding 1.0% of AKD into the pulp slurry is shown in Fig.l(b). Although the region for the peaks of the hydrocarbons from AKD overlaps that for the pyrolysates from cellulose, the eight major peaks of the ketones and AKDs are clearly observed in the pyrogram after a large number of polar pyrolysates such as levoglucosan, formed from cellulose, have been eluted. On the other hand, no significant peaks are observed in the latter part of the pyrogram of the paper containing no additives [Fig. l(c)]. Therefore, the AKDs and the ketones can be used as the key peaks for the determination of AKDs. Here, the AKD content in a given paper sample was calculated on the basis of the total intensities of the eight major peaks (1-8) observed in the pyrogram.Firstly, the relationship between the AKD content and the total inten- sities of the eight peaks relative to the sample mass was established by using the seven standard samples with AKD concentrations ranging from 0.004 to 2.3%. The correlation coefficient for the seven calibration points was 0.099. Then, the peak intensities in the pyrogram of a weighed unknown sample were correlated to the calibration graph thus prepared. The contents determined in this manner must be the total 150 rn 0 . .- .N 100 rn + 0, 0 E 50 85 1 ) 0 0.3 0.6 AKDs retained (%) Relationship between AKD content determined by pyrolysis- Fig. 3 GC and degree of sizing of paper samples. For details, see text ASA ( a ) n 0 10 20 30 40 50 Retention ti me/m i n Fig. 4 at 500 “C. (u) Pure ASA and (6) paper with 1.2% of ASA added Pyrograms of a paper sample containing ASA and pure ASA Table 3 ASA contents determined by pyrolysis-GC and degree of sizing of paper samples prepared in the presence of 0.12% of ASA Sample ASA in paper ASA retained ’ Stoeckigt degree No.(Yo) (Yo 1 of sizing/s 1 0.056 47 27.8 2 0.077 64 38.5 amounts of AKDs in the paper, including those AKDs that have reacted chemically with the hydroxy groups of cellulose and those that have been physically adsorbed, because the cellulose matrix is completely degraded to volatile products under the pyrolysis conditions used. This method was success- fully applied to the analysis of a sample containing only 0.001% of AKDs. The observed reproducibility was within 4% of the relative standard deviation for five repetitive runs with the same sample (sample No.3). As shown in Fig. 2, an almost linear relationship holds between the amounts of AKDs fed into the pulp slurry and the amounts retained in the paper as calculated by pyrolysis-GC. The broken line refers to the hypothetical relationship for complete retention (100%). The results obtained suggest that852 ANALYST, MAY 1992, VOL. 117 almost the same fraction (about 75%) of AKDs added is retained in the paper samples prepared by adding AKD at concentrations of between 0.025 and 1.0%. On the other hand, as shown in Fig. 3, the relationship between the AKD contents determined by pyrolysis-GC and the degree of sizing of the paper samples does not exhibit a simple linear tendency. The degree of sizing is approximately proportional to the AKD content up to about 0.15%, whereas the slope of the graph rapidly deviates from linearity at higher contents.Although initially (<0.15% AKDs retained), both the chem- ical reaction to the AKDs with the hydroxy groups of cellulose (I) and their physical adsorption onto cellulose (11) might lead to an almost linear increase in the degree of sizing, the chemical reaction would eventually be fully completed at a certain amount of retained AKDs (about 0.15%) when the available hydroxy groups of cellulose on the surface of the paper matrix had been consumed. On the other hand, physical adsorption can still proceed at the higher concentrations of AKDs added, so that the degree of sizing might then be increased mostly by the physically adsorbed AKDs.The ratio of the increase in the degree of sizing in the first stage to that in the second stage (see Fig. 3) is about 1 : 5. This result suggests that the chemically reacted AKDs might be about four times as effective as the physically adsorbed AKDs for the develop- ment of sizing. Determination of Retained AS A The pyrograms at 500 "C for ASA and the paper to which 0.12% of ASA has been added are shown in Fig. 4. The main peaks observed in Fig. 4(a) are due to intact ASA together with minor pyrolysis products such as alkanes and alkenes formed from the alkyl groups of the ASA. As the ASA generally consists of a complex mixture of various homologues and isomers with alkyl groups of different chain length and/or structure, the resulting chromatogram is very complex.As the small peaks of vaporized ASA are also clearly observed in Fig. 4(6), the ASA content in the paper samples can also be calculated from the intensity in a similar manner to that for the AKDs. The ASA contents determined by pyrolysis-GC, and the Stoeckigt degrees of sizing for the two paper samples prepared from the same pulp slurry containing 0.12% of ASA, are shown in Table 3. However, the degree of sizing of the two paper samples is different, in spite of the fact that the initial ASA dose was the same in both instances. The amounts of ASA retained in the paper, as determined by pyrolysis-GC, are consistent with the degree of sizing of the paper. These results demonstrate that pyrolysis-GC can be used to deter- mine trace amounts of neutral sizing agents in paper. References 1 Roberts, J. C., and Garner, D. N.. Tappi J., 1985, 68, 118. 2 McCarthy, W. R., and Stratton, R. A., Tappi J., 1987,70, 117. 3 Pan, Y. I., Kuga, S., Usuda. M., and Kadoya, T., Tappi J.. 1985, 68, 98. 4 Dart, P. J.. and McCalley, D. V., Analysr. 1990, 115, 13. 5 Yano, T.. Ohtani, H., Tsuge, S., and Obokata, T.. Tappi J.. 1991, 74, 197. 6 JIS P 8122, Testing Method for Stoeckigt Sizing Degree of Paper, Japanese Industrial Standards Committee, Tokyo, 1976. 7 Ohtani, H., Kimura, T., andTsuge, S . , Anal. Sci., 1986,2.179. Paper 1/03409J Received July 8, 1991 Accepted November 18, 1991