T 0565/00 (Carbon dioxide in sizing processes/CANADIAN LIQUID AIR) 14-03-2003

Reasons for the Decision

1. Main request

1.1. Novelty of the subject-matter of Claim 1 has been contested in view of document E12 which is a scientific article entitled "On the mechanism of sizing with alkylketene dimers; Part 2. The kinetics of reaction between alkylketene dimers and cellulose".

1.2.1. The article is based on lab scale experimental results with bleached kraft pulp using either tap water or deionised water as the aqueous vehicle, wherein AKD was added to the aqueous stock and paper sheets are formed from this stock after sufficient contact time (30 s) between the AKD and the pulp (page 34, summary, second paragraph and right-hand column, last paragraph to page 35, left-hand column, third paragraph). It was found that the reaction between the AKD and the cellulose of the fibers essentially takes place if the paper sheets are almost dry after a dry or wet curing treatment (page 34, Summary, second paragraph and right-hand column, lines 14 to 22). The experiments were carried out at pH 4, 6, 8 and 10 (in particular Figures 1 to 8) and showed that the rate of reaction increased with the alkalinity, i.e with increasing pH, both after dry and wet curing, as well as when using tap and deionised water (page 36, left-hand column, lines 1 to 9, page 37, left-hand column, last sentence above Figure 5 and right-hand column, last sentence above Figure 7 to page 38, left-hand column, first line below Figure 8).

1.2.2. However, it is stated in document E12 that the reaction rate is slower in deionized water than in tap water and that the pseudo first order reaction rate constant K1 (min-1) given in Table 4 shows for the former a decreased reaction rate over the whole pH range. Indeed, Table 4 contains K1 values for tap water and deionised water at pH 4, 6, 8 and 10 which are throughout higher in tap water and it is noted "that CO2 is probably absorbed at pH 10 in tap water and at pH-values exceeding pH 8 in deionized water" or "that CO2 from air may have been absorbed during sheet handling" and that "HCO3- is a strong reaction catalyst" (page 38, left-hand column, below Figure 8 and Table 4). From the analysis in Table 1 (page 35) it is known that tap water already contains this strong catalyst HCO3-, typically in a concentration of 0.48 mmoles/l.

1.3. Thus, given this information, it is evident that the higher reaction rates given for tap water in Table 4 of document E12 are due to its content of HCO3- ions right from the start and that the increase of the K1 values, at pH 10 for tap water and above pH 8 in deionised water, is due to absorption of CO2 from the air at alkaline pH into the aqueous vehicle either of the pulp or of the wet sheets formed therefrom (see also point 1.5.5. below). Such absorption of CO2 occurs as a technical inevitability during the subsequent dissolution in the pulp of CO2 under the formation of H2CO3 (carbonic acid) which in turn will dissociate, depending on the pH, in the manner well-known to the skilled person as confirmed by the figure on carbonic acid dissociation filed by the Respondent, to produce HCO3- ions, the maximum concentration of which are obtainable at a pH value of between 7 and 9.

1.4. The Appellant's arguments in favour of the novelty of the claimed process over that disclosed in citation E12 were the following:

It was not known from document E12 to inject CO2 gas in order to produce HCO3- ions in a catalytic amount, i.e in a sufficient amount to have a catalysing effect on the reaction between AKD and cellulose, while staying at non-acidic pH conditions. In particular, it was argued that the amount of CO2 present in air was low as was its solubility in water and that the term "probably absorbed" in document E12 did not suggest the injection of CO2 into the pulp. The term was rather used to explain the discrepancies contained in Table 4 insofar as the ratios K1(tap water) : K1(deionized water) were inconsistent with each other at the various pH levels. Nor did document E12 indicate that CO2 absorbed from the air would increase the rate of the sizing reaction. The Appellant agreed that CO2 injection generates a maximum amount of HCO3- ions when the pH of the aqueous medium was maintained at between 7 and 9 but concluded that, as a consequence, below pH 7 and above pH 9 a catalytic amount of HCO3- ions was not adequately provided.

1.5.1. These arguments are not convincing since the process of Claim 1 does not call either for a particular method of "dissolving of carbon dioxide gas in the aqueous vehicle" such as actively introducing the CO2 gas (e.g. by injection), or for a particular "catalytic" amount, let alone for a maximum amount obtainable at pH 7 to 9. Considering that CO2 injection decreases the pH of an aqueous solution, as submitted by the Appellant, Claim 1 is not even limited to conditions where the pH remains non-acidic after dissolving the CO2 gas in the pulp.

1.5.2. Instead Claim 1 covers, in the Board's judgment, any process where paper is made from an aqueous pulp which is formed from cellulosic paper-forming fibers and water and contacted with AKD at non-acidic pH, i.e. above pH 7, and in which pulp HCO3- ions are provided by CO2 gas entering the aqueous medium, be it by active injection or by passive absorption from the air, to catalyse the reaction between AKD and cellulose or, in other words, to increase the rate, i.e. the velocity, of that reaction, without being limited to conditions where the pH remains non-acidic after CO2 addition.

1.5.3. Such a process was, however, already suggested in document E12 where it is stated that HCO3- ions, probably generated by absorption of CO2 from the air at a pH above 8, act as catalyst for the AKD sizing reaction. This finding is confirmed in Table 4 of document E12 showing an increased reaction rate at non-acidic pH (in particular above pH 8) even in deionised water (see above point 1.2.2).

1.5.4. The ratios K1(tap water) : K1(deionized water), calculated for whatever reason by the Appellant for the different pH levels (see above point 1.4), are not mentioned in document E12, let alone any discrepancies or inconsistencies in this respect. Whilst the differences in these values may be based on various reasons, such as e.g. the different content of ionic species in tap and deionized water, possibly corresponding to a different CO2 absorption equilibrium, and/or a different dissociation behaviour of H2CO3 at the different pH levels, it is nevertheless clear that document E12 in Table 4 in combination with the corresponding explanations suggests that the reaction rate increases at a pH above 8 even for deionised water due to CO2 which has probably been absorbed dissociating after dissolution into the strong HCO3- catalyst.

1.5.5. The Board agrees that the only explanation given in document E12 for the results in the said Table 4 is that they are "probably" caused by the absorption of CO2. It is only for that reason that the process of Claim 1 is not explicitly disclosed in document E12.

However, as conceded by the Appellant, CO2 is usually always present in the atmosphere. Therefore, to work in an CO2-free environment would have required particular measures to remove this gas from the air. No such measures were reported in relation to the experiments of document E12 which, therefore, have not been carried out in the absence of CO2. Otherwise the suggestion in document E12 that CO2 absorption might have occurred would make no sense. It is generally known in the art that CO2 is absorbed by an aqueous medium, in particular if the pH of the medium is high. Therefore, a skilled reader would infer from document E12 the implicit teaching that CO2 was actually absorbed and consequently dissolved in the aqueous vehicle. This conclusion of the skilled reader is supported by the findings in Table 4 of document E12 (see points 1.2.2 and 1.5.4) for which the Appellant did not provide any other plausible explanation.

1.6. The decision under appeal referred to the Enlarged Board's of Appeal opinion G 2/88 in support of the novelty of the claimed subject-matter. However, Claim 1 does not relate to a new use of a known compound to achieve an effect as does opinion G 2/88 (see reasons Nos. 5.1 and 7) but to a process for the manufacture of a product. This opinion is thus not applicable in the present case.

The Board holds, therefore, that no technical process feature apt to provide novelty over the process disclosed in document E12 is added simply by indicating, in accordance with an embodiment covered by present Claim 1, that it is actually this absorption of CO2 which enhances via dissociation into HCO3- ions the reaction rate in AKD sizing.

1.7. The Board therefore concludes that the process of Claim 1 is anticipated by the disclosure of document E12.

2. Auxiliary request

2.1. Amendments

By the amendments made to the claims of the auxiliary request, for which a basis can be found in the application as originally filed, the scope of protection has been limited. The requirement of Articles 123(2) and (3) EPC are, therefore met. The amended claims also comply with the requirements of Article 84 EPC. This not having been contested by the Respondent, no further comment on that matter is necessary.

2.2. Novelty

Novelty of the subject-matter of Claim 1 over the disclosure of document E12 has been acknowledged by the Respondent due the new technical features of how to add the CO2 to the aqueous vehicle, namely by injecting it into the pulp flowing at a velocity effective to produce turbulent mixing, and a residence time of the CO2 in the flowing pulp of at least 30 seconds.

The Board agrees that these features are not mentioned in document E12 or, in combination with the other features of Claim 1, in any of the other cited prior art documents. The Board is, therefore, satisfied that the subject-matter of Claim 1 is novel.

2.3. Inventive step

2.3.1. The patent in suit is concerned with the general technical problem in the field of AKD sizing of paper in neutral or alkaline conditions that the reaction rate between AKD and cellulose is slow (column 1, line 56 to column 2, line 1).

2.3.2. The parties disagreed on the issue of whether document E2 or document E12 should be considered as the closest prior art. Both documents belong to the same series of articles under the common heading "On the mechanism of sizing with alkylketene dimers". Document E2 relates as "Part 4" of this series to a problem related to that in the patent in suit, namely to "The effects of HCO3- ions and polymeric reaction accelerators on the rate of reaction between alkylketene dimers and cellulose". It has been found that from the various additives investigated (page 41, Table 2) in this respect, HCO3- and cationic basic amines were the most efficient sizing accelerators (page 42, left-hand column, fourth paragraph). Document E12, as mentioned under 1.1 above, is part 2 of the same series and concerns the kinetics of the AKD sizing reaction. It includes, however, the issues concerning the rate of reaction and the catalytic influence of HCO3- ions (Summary, page 39, left-hand column, third full paragraph and Table 7). The Board holds, therefore, that both documents are equally suitable as a starting point for assessing inventive step.

2.3.3. Technical problem and solution

According to documents E2 and E12 the HCO3- ions are preferably provided by adding NaHCO3 (document E2, Table 2, document E12, Table 7).

Whilst the Respondent did not accept the Appellant's argument that the introduction of Na+ ions might be a problem in the sizing process, the parties nevertheless agreed insofar as the technical problem solved by the claimed process in view of such prior art can be seen in avoiding the using of NaHCO3 for HCO3- generation for one reason or another. The Board also follows this approach, the more so as it is evident that this technical problem is solved by injecting CO2 gas into the aqueous pulp under conditions which provide effective mixing and a residence time sufficient for the CO2 to be dissolved in the pulp. The Board considers it credible that this can be achieved by injecting the CO2 gas at a point where the liquid velocity is high enough to produce turbulent mixing and which allows a hydraulic residence time of the CO2 in the flowing stream of at least 30 seconds in accordance with the claimed process.

2.3.4. It remains to be decided whether, in view of the cited prior art documents, it was obvious for someone skilled in the art to solve this technical problem by the above means.

2.3.5. It follows from the above (see in particular 1.2 and 1.5) that the experiments disclosed in document E12 which have been carried out in a laboratory suggest CO2 gas absorbed from the air as being a possible source for HCO3- generation. Whilst it is true, as indicated by the Appellant, that injection of CO2 gas is not suggested in document E12, it is, nevertheless, obvious for those skilled in the art that it is not technically feasible in a large scale process simply to wait and expect that the gas be passively absorbed in amounts sufficient to accelerate the AKD sizing reaction to the same extent as would be possible by adding NaHCO3. On the other hand, the technical possibilities for introducing a gaseous compound into a liquid in a practicable manner are limited and the Appellant has not provided any evidence that a skilled person would select any other methods than absorption and injection. Thus, the Board agrees with the Respondent that for those skilled in the art injection is at least the most convenient way of introducing the CO2 gas into the pulp on a large scale.

2.3.6. The Appellant did not comment on the effects provided by the features relating to the turbulent mixing and the residence time of at least 30 seconds of the CO2 gas in the pulp. The patent in suit merely mentions in this respect that the point of injection is located downstream of a pump where the velocity of the pulp is sufficient to provide turbulent mixing in a feed line which is long enough to provide "in this way adequate dissolution of the carbon dioxide in the flowing stream is achieved" (column 4, lines 45 to 49). Certainly, this does not necessarily mean that this way is the best one or the only one for mixing and dissolving the gas "adequately".

In the Board's opinion, it is, however, within the general knowledge of those skilled in the art that a reaction catalyst should be evenly distributed in the reaction medium to provide homogenous reaction and that a retention time is needed for dissolution of the CO2 in the aqueous vehicle, as a prerequisite for the generation of HCO3- ions (see 1.3). Therefore, the skilled person is in any event required to take such steps.

The Appellant did not provide evidence as to any unexpected effect such as a particular advantage of the claimed injection at a point where the velocity of the liquid produces turbulent mixing and which allows a hydraulic residence time of the CO2 in the flowing stream of at least 30 seconds. Further, it is apparent from document E8 that there existed no prejudice in the technical field of paper making against doing so since it discloses exactly the same procedure of injecting CO2 under turbulent conditions at the discharge side of a pump in order to effect acidification, or by implication dissolution of the CO2, in the subsequent line (column 2, lines 47 to 54 and Figure).

The features related to mixing and residence time are, therefore, considered to be one option amongst others which a skilled person would select according to circumstances in order to find adequate conditions to generate HCO3- ions from CO2 gas in an industrial process.

2.4. Therefore, the Board concludes that, for the purpose of avoiding NaHCO3 addition as a source for HCO3- generation as disclosed in documents E2 and E12, the skilled person, following a suggestion in document E12 to generate it from CO2, would have tried to produce the HCO3- ions efficiently by injecting CO2 under conditions known to produce effective mixing and dissolution in the pulp.

Consequently the subject-matter of Claim 1 lacks an inventive step and does not meet the requirements of Article 56 EPC.