8.4 Consideration
164 The advance in the art disclosed in the specification is the discovery that certain silane based small molecules have been found to be effective in treating aluminosilicate scale in a Bayer process stream (page 1 lines 4 - 7). The background to the invention describes the problems associated with DSP scale in the Bayer process and highlights the prior art solution that was based on the addition of polymer materials containing three alkyloxy groups bonded to one silicon atom (page 3 lines 8 - 10). That solution had problems with unwanted degrees of viscosity, which made handling and dispersion (miscibility) of the polymer through the Bayer liquor problematic (page 3 lines 16 - 18). The specification notes that it is an object to address or ameliorate these problems, which is said to have been done in several aspects of the invention.
165 Each aspect of the invention then described is prefaced with a reference to it being a method for the reduction of aluminosilicate containing scale in a Bayer process. There is no dispute that this process is conducted at an industrial, refinery level. The focus of each aspect of the invention is on the synthesis of the at least one small molecule comprising at least three components, described most generally by reference to the Markush diagram:
where R1, R2 and R3 are selected from an extremely broad group of constituents and are bonded to a nitrogen atom. Later aspects and embodiments narrow the selection.
166 The detailed description states that it was "discovered that dosing of various types of silane-based products can reduce the amount of DSP scale formed" (page 6 lines 21 - 22). The embodiments described in the detailed description again focus on the reaction product to be produced. To reinforce the point, the specification says that although some of the small molecules have been identified in the prior art in relation to other unrelated applications, "their effectiveness in reducing Bayer Process scale was wholly unexpected" because the prior art taught that only high weight polymers are effective (page 32 line 10 - page 33 line 17). The specification then sets out the three advantages said to be conferred by using small molecule based inhibitors as opposed to polymeric inhibitors (page 33 lines 18 - 26).
167 It is in this context that the specification sets out the two examples which are said to be presented "for the purposes of illustration and are not intended to limit the scope of the invention" (page 34 lines 3 - 5). I have set out much of the content of the examples in section 4.1 above.
168 Example 1 identifies a semi-batch method as being preferred for performing the synthesis of A, G and E when compared with a batch method (both of which were known to those skilled in the art).
169 The primer records that the general concept of "batch", "batch-on-batch" and "semi-batch" methods is well known among chemists. In summary:
(1) A "batch" method is a method where all reactants are reacted together with one another at the same time.
(2) A "batch-on-batch" method is a method where all reactants are reacted together with one another, but the reaction is split into multiple "batch" reactions (e.g. 10% of the total amount of the reactants is reacted in a first "batch" reaction; then the next 10% of the total amount of the reactants is reacted in a second "batch" reaction, and so on). A "batch-on-batch" method may be a series of smaller scale batch processes, to reduce the scale of each reaction (e.g. for safety reasons in an industrial manufacturing process, or to manage the amount of heat produced by a reaction that is exothermic), with all of the reaction products ultimately combined.
(3) A "semi-batch method" is a method where all reactants eventually end up in the same reactor, but the order of addition (including the rate, timing, amounts and proportions of the different reactants present) is varied. The purpose of the variations in a "semi-batch method" is to control the reaction conditions during the course of the reaction, particularly where the reaction is exothermic.
170 The semi-batch method is said in the patent application to allow for a much better control over the overall reaction by adding only a portion of the G and E in a form of a slow feed to initiate the primary epoxide ring opening reaction followed by the slow continuous feeding of G and E (page 35 lines 1 - 4). The evidence indicates that the person skilled in the art would understand that A is added first before the feed of G and E is commenced. The semi-batch method is said to provide a better overall yield of the active compounds in the product, also avoiding undesired side reactions (page 35 lines 4 - 6). It is also said to reduce the amount of methanol that is generated and isolated during the reaction and to produce a reaction mixture with a lower viscosity (page 35 lines 12 - 20). Dr Power observes that Table 1 records that the reaction temperature of the semi-batch method (as opposed to the batch and batch-on-batch methods) was more controlled, with very little methane involved, which suggests that the methoxysilane groups remained mostly intact. The end product also did not show the same degree of coupling-related viscosity of the batches produced using the batch and batch-on-batch methods.
171 Example 2 is said to provide examples of the relative DSP scale inhibition of various A:G:E small molecules formed during the synthesis reaction "disclosed above". Although Professor Easton in his written evidence was unsure which of the three methods described in Example 1 was used, by the hearing Cytec agreed that it was the semi-batch method. The example is of a laboratory test that involved adding a small amount of sodium silicate to a Bayer liquor to increase the super saturation of silica in solution, dosing the liquor with the scale inhibitor being tested, raising the temperature to 105°C (just below the boiling point of a Bayer liquor) and heating it for a few hours, and then cooling the samples and measuring the amount of DSP scale formed. The results illustrate that of the five amines tested (in conjunction with GPS and E) as the scale inhibitor, hexane diamine performed the best at high and low doses, the next best was 1-amino-2-propanol and the third best was ethylene diamine. These are the three amines claimed in claim 1.
172 Having regard to the disclosure of the specification and the claims, the embodiment which is described and around which the claims are drawn is the use of a product mixture containing small silane-based small molecules synthesised from the reaction of components A (either H, ED or AP), GPS and E, which mixture is to be added to Bayer process stream in an aluminosilicate scale inhibiting amount. The method is directed to the application of the synthesised product mixture in a Bayer process stream which is on an industrial scale. The invention described and claimed resides in the discovery that the use of small molecules of the type described and claimed can be useful in the inhibition of scale in the Bayer process. It is not an invention for a method of industrial (or indeed laboratory) methods of synthesis.
173 For the following reasons I am not persuaded that further provision of specific information going to the manner in which the semi-batch process outlined in example 1, and more particularly the details of the synthesis reaction used to achieve the HGE product in the laboratory in example 2, was necessary in order for Nalco to satisfy the best method requirement of s 40(2)(aa) of the Patents Act.
174 First, at a general level, it must be noted that examples 1 and 2 concern laboratory tests performed by Nalco. There is no suggestion that, by the time of the filing of the complete specification, Nalco had produced any industrial-scale batches of product mixtures that were suitable for use in accordance with the method of the claims. This provides an important distinction to the circumstances in Servier, where the evidence indicated that the inventors had developed a method of making something that did fall within the claims. The focus of the debate in the present case is one step removed from the invention claimed in the patent application, and concerns the extent to which the outcome of laboratory testing might or would be useful to the person skilled in the art if they wished to make something that could be applied within the method of the claims.
175 Secondly, examples 1 and 2 taught that the inventors had learned that the use of a semi-batch method of production was preferable, that using different amines would give rise to different efficacies of DSP inhibition and that the HGE product was the most effective of the mixtures that the inventors had tested. The semi-batch method is a standard process known in the field that, the specification notes, allows for better control of the overall reaction and the reaction temperature and provides a better overall yield of the active compounds in the products, also avoiding undesired side reactions that would result from using a batch process. In such a process, the temperature control may be achieved by adjusting the feed rate of the reactants as well as the cooling mechanisms used, which will vary from plant to plant depending on the available equipment. The specification provides information that controlling the temperature of the reaction is important in order to achieve the optimal yield of small molecules. Professor Easton considered that it also provides a number of temperature parameters for the reaction, being an initial temperature of 65-70°C to initiate the reaction and an optimal temperature range of between 82-93°C. It also teaches that a prolonged temperature above 120°C is to be avoided in order to avoid unwanted side reactions.
176 I accept that once that information has been provided, it would be a matter of routine for the skilled process engineer to design an industrial synthesis process according to the equipment available. Viewed in this light, Figure 2 in the patent application provides an indication of the preferred general approach that should be taken to synthesising the reaction product mixture. This, of course, is not a complete answer to an allegation of a failure to disclose the best method (see Servier at [135]) but it goes some way towards that outcome, depending on the materiality of facts held by the inventor not disclosed in the specification.
177 Thirdly, typically, a scale inhibiting amount of a composition of the type claimed in the patent application (and also the prior art Max HT product) would be dosed continuously into the Bayer stream at a rate of between 10-50 kg/hr, amounting to hundreds of tonnes each year per refinery. That is a radically different amount of the composition to that produced in a laboratory scale synthesis.
178 The evidence demonstrates that laboratory-scale methods of synthesis can only provide a rough starting point for the development of a method of industrial scale production of synthesised reaction mixtures in quantities necessary for industrial application. I am not persuaded that the provision of more granular information detailing the way in which the HGE product was made in the laboratory, including the various parameters of the semi-batch synthesis method used, would provide information of sufficient use to the person skilled in the art seeking to perform the invention. This is because the skilled addressee would, in any event, have to develop an industrial-scale production method of synthesis suitable for synthesising the small molecule product mixture for industrial application. As I have noted, there is no suggestion that Nalco had performed any industrial scale production of the reaction product mixture described.
179 Mr Bellwood gave evidence that to work the synthesis reaction on an industrial scale he would put the amine into the reactor, bring it up to the initiation temperature and then, very quickly, feed in an amount of the epoxide (that is, GPS and E) which he would have established was enough to "kick off" the reaction. He would then, once the temperate in the exotherm started to kick off, commence a continuous feed of GPS and E to maintain that reaction. At the same time he would also apply coolant to the vessel in order to extract the heat, the aim being to balance the heat generated by the exotherm with the heat extraction system of the reactor, to maintain as close to a constant temperature as possible. He would be "looking constantly to balance the rate of reaction through the feed rate of the G and E" and then continue that to completion when all of the reactants have been added.
180 Dr Power did not disagree with that approach. He said in response:
DR POWER: I don't disagree; it's one way of doing it, absolutely, but there's any number of other ways, and the chemical engineering of it would be a whole thing in itself. You might want to use a series of continuous stir tank reactors if you were wanting to produce a lot of this stuff in a commodity way. It would be unlikely that if you were wanting to produce large quantities of it that you would use batch reactors. It probably is continuous to tank reactors, and then that would - and I also would say that, having all the A there at the start is one assumption about how the chemistry would work better. As I pointed out before, there's other ways of doing that as well, so there's any number of ways of doing this, and then the chemical engineering of it is a whole - another set of developments. So the patent doesn't give you any guidance at all on any of that. This is just a way of producing a sample in the lab. That's all it is.
(Emphasis added)
181 Dr Audet gave evidence to similar effect.
182 These factors persuasively illustrate that the parameters of the semi-batch synthesis reaction used to produce the best performing reaction product mixture in examples 1 and 2 of the patent application could not in any event simply be scaled-up for industrial scale production. The skilled industrial chemist must design the industrial scale reaction having regard to the plant equipment used, including its size and cooling capacity, which will affect the feed rate of the reactants necessary to maintain the temperatures that will provide an optimal yield. This is a balancing exercise. I accept that laboratory scale reactions such as those in example 1 can do little more than provide general guidance as to the preferred approach to be adopted on an industrial scale. In line with this, example 1 guides the skilled addressee to the optimal reaction conditions known to the applicant based on its laboratory testing. I am not satisfied that the supply of more specific details of the process used by the applicant in the examples would be of practical assistance to the person skilled in the art.
183 Fourthly, there are relevant differences between individual Bayer liquors such that the results provided in the examples are unlikely to have a material bearing on a final process of synthesis of the reaction product for use on an industrial scale. In this regard the following background matters may be noted:
(1) All Bayer liquors are aqueous solutions of sodium aluminate which inevitably contain varying amounts of inorganic impurities including sodium carbonate, sodium sulphate, sodium chloride, sodium silicate, and a complex mixture of organic impurities.
(2) There are significant variations between individual refineries in all of these parameters, particularly organic impurities.
(3) The amounts of suspended solids in Bayer liquors varies between Bayer liquors, including the levels of suspended solids in the form of red mud.
(4) The types of aluminosilicate scale differ between Bayer liquors and include sodalite, cancrinite and vishnevite.
(5) The tendency and type of aluminate scale to form at a particular refinery will depend on a wide variety of factors including the type of bauxite, which will in turn determine the process conditions.
(6) Variables such as these mean that some Bayer refineries perform digestion at a lower temperature range than others: where the bauxite is predominantly gibbsite, the temperature is at least 140-145°C, and where it is boehmite, the temperature is at least 250°C.
184 Mr Bellwood gave evidence that the composition of the Bayer liquor is something that has a significant bearing on the fine detail of the product mixture which is most efficient at a particular refinery. It is his experience that the composition of the reagent, relative to its chemical family, will need to be optimised in relation to the varying properties and composition of the Bayer liquor. He anticipated that in applying the method of the patent application, the molar ratios (i.e. the relative proportions) of reactants would need to be adjusted to optimise the composition for specific refineries. In this regard, whilst the overall chemistry approach is likely to be the same for all refineries, specific parameters in the synthesis, including molar ratios, are routinely and necessarily adjusted as part of the process to optimise for product performance. Dr Audet gave evidence to similar effect.
185 Dr Power agreed that varying the molar ratios of the reactants may cause the product mixture to be more or less effective at inhibiting DSP scale depending on the temperature of the Bayer liquor before it goes into the heat exchangers in a particular Bayer refinery and also the pH of the Bayer liquor.
186 Dr Power considered that the similarities between Bayer liquors, including high ionic strength, strong alkalinity, and high levels of alumina and silica, to be more important than their differences when it comes to the design of scale inhibitors. However, his evidence in cross-examination in relation to the Max HT product (and its subsequent iterations Max HT 500 and 550) served to demonstrate that: (1) the performance of Max HT varied across different Bayer liquors such that it was necessary to adjust its dosage from refinery to refinery to achieve scale inhibition efficacy; (2) Max HT was developed over time to become more tolerant to the temperature and pH of Bayer liquors, as well as to perform better in the presence of suspended solids, factors which vary between refineries; and (3) in principle, one option open to an additive manufacturer who wished to improve the efficacy of an additive would be to vary the molar ratio of the reactants to optimise it for a particular Bayer liquor. In my view, this evidence tends to support the position advanced by Nalco that the synthesis of the reaction product mixture will be informed by the particular characteristics of the Bayer liquor in which it is intended to be used. Certainly it serves to demonstrate that testing on one Bayer liquor in a laboratory could not be regarded as particularly determinative of the final form of a synthesised product mixture useful in any given industrial operation.
187 Furthermore, the product mixture produced from the reaction envisaged in claim 1 has the potential to include a greater or lesser amount of hydrophobic constituents. There is some uncertainty as to how those hydrophobes would interact with suspended organic solids in a particular Bayer liquor, either by adsorbing to those solids or instead interacting, as intended, with the silane groups of the aluminosilicate seed crystals. Depending on that interaction, the scale inhibiting aspect of the reaction mixture would be more or less effective. Dr Power accepted that it was a matter of speculation as to how that would be revealed in terms of efficacy. Professor Easton accepted that to determine the effect of a greater number of hydrophobes on a molecule against its tendency to adsorb to solids would require testing with the Bayer liquor in question. Such speculation serves to demonstrate the need for testing of any product mixture in the Bayer liquor in which it is intended to be used.
188 Fifthly, based on evidence given by Professor Easton, Cytec contended in closing submissions that the examples ought to have included, in relation to the semi-batch synthesis method used to produce the HGE product, the following details:
(1) the order and sequence of the addition of the reactants;
(2) the rate of addition (feed rate) and mixing of each of the reactants being added to the reaction;
(3) the amount and relative ratio of the reactants;
(4) the temperature controls used for the reaction; and
(5) the reaction time.
189 In dissecting the examples in this way, Cytec tends to distract attention from the correct approach. As noted, the invention disclosed and claimed is not to a method of industrial or laboratory synthesis of the identified small molecules but to the novel application of the reaction product mixture on an industrial scale in a Bayer process stream.
190 Sixthly, and further to the fifth point, the evidence adduced by Cytec does not serve to demonstrate that there is a relevant correlation between, on the one hand, laboratory test results of the type it alleges are omitted from the patent application and, on the other, industrial scale process development. Indeed, Professor Easton was not asked in his evidence in chief to address the utility of the information in the examples for a subsequent industrial application, which might have been more material to the present question. His evidence focussed upon whether he would be able to replicate the best results set out in the examples. It was upon this premise that the five omitted items of information now relied upon by Cytec are based. However, as I have noted, I am not satisfied that the laboratory experiments exemplified in the examples of the patent application would provide a useful proxy for industrial manufacture in any event.
191 Finally, I turn specifically to the five points made by Cytec set out above at [188].
192 The first of these (order and sequence of addition of reactants) is, as noted earlier, addressed in example 1. The evidence of the Bayer process experts was that in light of the disclosure in the specification, they would understand it to teach to add the amine to the reaction vessel first, then add a small amount of GPS + E to kick off the reaction, and then feed in GPS + E together to maintain the reaction. Although the language used in the example is not completely clear, I accept that this is how the person skilled in the art would understand the teaching.
193 In relation to the second (feed rate and mixing of reactants), the patent application discloses that using the semi-batch process is recommended as it enables the feed rate of the reactants to be controlled, which controls the exothermic reaction. For reasons I have explained in more detail above, the specific feed rate and mixing used for the HGE product is not material to the disclosure when it is applied on an industrial scale.
194 In relation to the third (the amount and relative ratio of the reactants), each of the Bayer process experts agreed that if seeking to implement the invention in an industrial setting they would commence with a 1:1:1 molar ratio of the reactants and test the reaction product mixture for efficacy. They would do this even if example 2 had provided specific or different molar ratios for the HGE product. I have noted above that Dr Audet and Mr Bellwood explained that they would do this because the molar ratios should be optimised for the particular Bayer liquor used in each refinery. I accept this evidence.
195 In relation to the fourth (the temperature controls used for the reaction), the patent application discloses: (a) that the epoxide ring opening reaction between the A reactant and the G and E reactants is initiated at a temperature of approximately 65-70°C; (b) prolonged exposure at high temperatures above 120°C is to be avoided so as to maximise the yield of the small molecules; (c) the temperature of the reaction must be controlled to avoid undesired side reactions and to provide a reaction mixture with lower viscosity (as exemplified by Table 1); (d) the semi-batch synthesis method is recommended as it enables the feed rate of the reactants to be controlled, limiting the rate of the exothermic reaction, and ensuring that the temperature is controlled within the optimal range of approximately 82-93°C (as exemplified by Figure 2).
196 In relation to the fifth (reaction time), the Bayer process experts agreed that the time it will take for the reaction to complete will differ depending on the amount of the reactants used, the temperature at which the reaction is performed, the production facility used, and the feed rates.
197 Having regard to the matters addressed above, in my view, Cytec has not demonstrated that the patent applicant has failed to disclose the best method known to it of performing the invention within s 40(2)(aa).
198 I should note that in cross-examination, Cytec suggested to Dr Audet that if he were to have been provided with a "recipe book" containing all of the details used to synthesise the HGE product as at the priority date, and he were asked by a commodity chemical company to develop the best possible Bayer process antiscalant product using the patent application, he would "take it without hesitation", and indeed pay for it if he was short on time. Dr Audet accepted these propositions, qualifying his acceptance by pointing out that he would buy the recipe book if he wanted a "guaranteed solution", because without the recipe book, he could not be so guaranteed.
199 Understandably, Cytec placed reliance on this evidence in its closing submissions. However, the difficulty with this exchange is that it appears to be inconsistent with Dr Audet's more detailed and reasoned evidence that he would need to optimise the synthesis parameters of the reaction product mixture for industrial sized manufacture, as well as the individual Bayer liquor in each refinery in which the reaction product mixture is to be used. I do not think, by this evidence, Dr Audet intended to recant from his earlier views, and it was not suggested to him that he did. I take this evidence to indicate that such a recipe book could assist the skilled addressee in developing a reaction product mixture, on a laboratory scale, for the particular Bayer liquor used in example 2. However, this is not the invention claimed in the patent.
200 In a similar vein, Mr Bellwood accepted that whilst that is not an approach that he would take, having the recipe book would provide him with a "starting point" for the optimisation process that he would undertake in any event. Mr Bellwood accepted that having the recipe book would mean that he could make a reaction product mixture that had the scale reduction qualities of the HGE product. He also accepted that he could perform tests on other Bayer liquors to confirm the efficacy of the mixture in the liquor of those refineries. If satisfactory, he could take the product forward to plant testing and scaling up. By these answers Mr Bellwood accepted that having more information of this type could be useful. However, again, I am not satisfied that by this short passage of answers that Mr Bellwood should be understood to have retreated from the more reasoned explanations to which I have referred above about the distinction between the outcome of laboratory tests and the method ultimately likely to be used for industrial scale synthesis. Most scientists offered more information would accept it, but that fact does not of itself mean that there has been a failure to supply a best method.
201 Accordingly, the challenge advanced on the basis of s 40(2)(aa) of the Patents Act fails.