England and Wales High Court (Patents Court) Decisions

Before: MR. DAVID YOUNG Q.C.

Sitting as a Deputy Judge of the High Court

IN THE MATTER OF the Patents Act 1977

and

IN THE MATTER OF European Patents (UK)

Nos. 0 569 114; 0 020 907; and 0 557 835;

each granted to Th. Goldschmidt AG.

AND

BETWEEN:

TH GOLDSCHMIDT AG and another

-and-

Defendant/Petitioner

EOC BELGIUM NV and others

Mr. Christopher Floyd Q.C. and Mr. Richard Arnold instructed

by Bristows for the Claimant

Mr. Michael Fysh Q.C. and Mr. Michael Tappin instructed

by Walker Morris for the Defendant

Hearing dates:

8th, 9th, 10th, 13th, 14th, 15th, 16th December 1999

1. On the 30th July '98 Mr. Justice Pumfrey ordered the above Petition and Action and Counterclaim to be tried together.

2. In addition to the foregoing there is an application to amend European Patent No. 0,020,907 which is unopposed.

3. For convenience unless otherwise indicated, in this Judgment I will refer to the Claimants (and Defendants in the Counterclaim) as Goldschmidt and the Petitioners, who are Defendants in the action and Claimants in the Counterclaim as EOC.

4. The subject matter of the Goldschmidt patents relates to improvements in the method of preparation and formulation of the known amidobetaine co-surfactants which have been used in personal care products such as shampoos since the 1960s and since the mid '80s in hand dishwashing liquids.

5. Amidobetaines are produced in a two step process. In Step 1 a fatty acid (or fatty acid ester) such as a naturally occurring saturated or unsaturated fatty acid is reacted with an amine such as dimethylamino propylamine to form an amidoamine (or aminoamide)

6. In Step 2 the amidoamine is reacted with sodium chloroacetate or chloroacetic acid to form the amidobetaine in a quaternisation reaction:

7. The three patents in suit relate to improvements in the quality of the amidobetaine prepared according to the reaction the subject of Step 2 and I shall consider each patent seriatim.

EUROPEAN PATENT 0,020,907 (THE pH PATENT)

8. The pH patent has a priority date of 30th June 1979 and addresses the problem of unreacted amidoamine and chloroacetic acid in the final amidobetaine product. Amidoamine was known to be an irritant or have a sensitising effect.

9. At page 3 it is stated that it is essential to the invention that the pH of the alkaline solution must have a value of from 7.5 to 10.5 measured at 98^oC over the entire course of the reaction i.e. in particular also toward the end of the reaction. This is to be contrasted with the prior art processes which customarily operated essentially at neutral pH.

10. Reference is made to DE-B-2063422 (Koebner) in which it is stated that a conversion of 90% is the maximum that can be achieved. Koebner's solution to the problem of removing the tertiary amine impurity was to convert the same into the corresponding N-Oxides using peroxides.

11. At page 3 lines 23 to 34 it is stated "if the process is carried out within the pH range of from 7.5 to 10.5 it is not possible to detect any fatty acid amide dialkylamine by thin-layer chromatography after a reaction time of about 3 hours, the detection limit being 0.02% by weight.

12. A pH of 8 to 10 is to be particularly preferred, since above a pH of about 8 the excess unreacted alkali metal salt of the w -haloalkyl carboxylic acid decomposes. If it is desired to ensure that no organically bonded chlorine is detectable in the reaction product, a reaction time of about 8 to 10 hours is recommended".

13. The patent goes on to explain that above a pH of 10.5 significant decomposition of the desired betaine will occur whilst below a pH of 7.5 the degree of conversion and reaction progress deteriorates.

14. At page 4 the patent states that in contrast to Koebner the invention "attempts … to allow the reaction to proceed quantitatively without the formation of product impurity.

15. The reaction product obtained according to the invention thus consists virtually only of the desired active substance and not of a mixture of active substance and unreacted impurity".

16. Claim 1 is for a process for the preparation of amidobetaines

"which comprises carrying out the quaternisation reaction, over its entire course, in alkaline solution which has a pH of from 7.5 to 10.5 measured at 98^oC"

17. Claim 2 is to the process as claimed in claim 1.

18. The Claimants have sought leave to amend claim 1 to a process for the preparation of amidobetaines

"characterised in that the reaction mixture is heated to 98^oC and in that the quaternisation reaction, over its entire course, is effected in an alkaline solution which has a pH value of 7.5 to 10.5 measured at 98^oC."

(underlined words are the subject of the amendment). Such amendment is unopposed and prima facie is allowable and accordingly I shall consider the claim as amended.

19. The principal issue on construction of claim 1 is whether the words "over its entire course" refer back to the reaction mixture being heated to 98^oC and if so does the temperature of 98^oC mean it must be maintained at that precise value.

20. The significance of 98^oC is that it is said to be the highest temperature the reaction mixture can be heated to under atmospheric pressure without undue frothing caused by the surfactant (the Capuccino effect).

21. Dr. Jurges a retired chemist who worked for Hoechst in their surfactant division (First Report, para 28) refers to 98^oC as the standard temperature for the quaternisation reaction being just below the boiling point.

22. Professor Grigg an expert for the Defendants (Supplementary Report, para 2) agreed that heating to 98^oC meant heating just below the boiling point.

23. The quaternisation reaction is an exothemic one and has to be controlled. I do not construe the specified temperature of 98^oC as being anything other than a requirement to heat to and maintain the temperature just below the boiling point.

Dr. Jurges (Day 1/107¹¹-108²¹ and Day 2/226) believed a range of 90-98^oC would be a practical range in which to operate. He would not expect there to be a difference in the product whether it was heated at 90^oC or 98^oC nor in the process other than to allow it be operated more easily if it was allowed to drift.

24. Mr. Fysh Q.C. for the Defendants also contended that the words "over its entire course" meant that the reaction has to continue until the level of amidoamine present in the reaction vessel is below the levels of the prior art indicated in the patent (namely less than 0.3 -3%). I do not accept one can read such a limitation into the claim. Whilst the patent promises that the process is capable of achieving such levels of purity by the use of the process claimed, it also states that the process can be distinguished over the prior art (Koebner) in that Koebner's quaternisation reaction will only produce a 90% conversion and to achieve a higher level of purity it was considered necessary to carry out an after treatment. I consider the "course of the reaction" is dictated by the operator as to whatever level of purity he is seeking.

25. The following three products are alleged to infringe:

(1) EUROQUAT LA/SURFAC LA: (referred to as LA)

(2) EUROQUAT C45/SURFAC B45/DEV112A (referred to as C45)

(3) EUROQUAT C47/SURFAC B45 (referred to as C47)

26. All three products are made by EOC Belgium and distributed in the UK by Surfachem. EOC admits it is liable for the acts of Surfachem. C45 has not been made since March 1997 whilst C47 has only been produced since May 1997.

27. Mr. Delahaye, general manager of EOC Surfactants NV describes the preparation of each of these products which I shall now consider in turn:

LA: This is a product with low levels of solids (circa 35%) and is only alleged to infringe the pH patent.

28. There is no dispute that the quaternisation reaction is carried out over the course of the reaction at a pH between 8 to 10 (9.0 ⁺ 0.5) measured at 98^oC. The reaction is a batch operation and there are variations from batch to batch. Data relating to the production of batches supplied to Surfachem is set out in Exhibit PD7 which has been extracted from EOC's batch production sheets.

29. The only question is whether the reaction mixture is heated to 98^oC in the manner required by the amended claim 1.

Mr. Delahaye (Day 4/532²⁴-533³³) stated that the aim was to remain between 90-100^oC. Three of the batches (18, 44 and 178) supplied to Surfachem have recorded temperatures of 99^o, 97^o and 100^o respectively whilst the other batches (43, 172, 173, 176 and 244) have lower recorded temperatures of 90-93^oC.

30. Such recorded temperatures were taken at various specified times. However there is no continuous temperature recording.

31. I consider that as regards the first three mentioned batches, they were heated to 98^oC and the evidence of Mr. Delahaye is they would have been maintained between 90-100^oC and so there is no doubt the production of such batches infringe both claims 1 and 2 of the pH patent.

32. As regards the other batches there is no evidence they reached 98^oC but I do not consider the precise temperature of 98^oC to be critical to the working of the invention. Rather the significance of 98^oC is that it is the maximum temperature one can safely operate below the boiling point at atmospheric pressure without frothing. Working between 90^oC-95^oC would I believe constitute an obviously immaterial mode of operation which as a practical matter could not have been intended to have been excluded by the patentee (see "Improver Questions" (1990) FSR 181 at 189).

C45: This is a product with high levels of solids (circa 45%). Production was discontinued in March 1997. Apart from various samples, designated A, B, C, X and Y, only one commercial batch of C45 was supplied to Surfachem (February 1997). No precise records of the production of this batch can be identified - indeed the batch supplied to Surfachem may well be a blend of a number of production runs which had been conveyed to a bulk storage tank prior to its supply to Surfachem.

33. Exhibit PD8 tabulates the pH and recorded temperature data of production runs made 1-3 weeks prior to the supply of such samples.

34. It is accepted that the pH of such production runs falls within the 8 to 10 range.

35. As regards temperature some of the runs exceeded 100^oC due to the fact that they were performed in a reactor that can be put under pressure. Thus for Sample B runs 45, 49 and 74 were heated to temperatures of 106^oC-109^oC whilst runs 53, 56, 59, 63, 67 and 71 record temperatures of 92-98^oC.

36. There is no suggestion in the pH patent to heat above "the customary reaction temperature" of 98^oC (page 3 lines 19-22). Indeed part of the Claimants' defence to the validity of the EP 0 557 835 (the temperature patent) is that in this art as a matter of choice one would not use autoclave or other pressure vessels and that the amidobetaine and the amidoamine might degrade at higher temperatures (see Jurges Report para 94). Hence temperatures above 100^oC cannot be considered to be obviously immaterial variants of the process claimed in the pH patent.

37. It follows that so far as samples A, B, C, X and Y were made up of amidobetaine product derived from production runs conducted at temperatures between 90-100^oC, to such an extent such samples contain product which falls within claims 1 and 2 of the pH patent.

C.47. This is also a product with high levels of solids. Production was commenced in May 1997 and pH and temperature data relating to the two deliveries to the UK form the subject of Exhibit PD8.

38. It is common ground that the infringement issues for C47 are the same as for C45.

39. Thus for C47-Com 1, amidobetaine product from Runs 78, 81, 84 and 87 fall within claims 1 and 2 whilst product from Runs 90, 94 and 98 (at recorded temperatures of 109-112^oC) does not. Likewise for C47-Com 2 product from Runs 112, 120, 123, 129, 133, 137, 140 and 143 falls within claims 1 and 2 whilst product from Runs 101, 105, 109, 114, 126 and 147 does not .

40. By 1979 it was common general knowledge that residual amidoamine had an irritating or sensitising effect and that anyone interested in producing amidobetaines would wish to minimise the presence of unreacted amine by conversion of the amine to the betaine.

41. Furthermore it was common knowledge to a chemist interested in the quaternisation reaction that the rate of conversion of the amine to the betaine is temperature dependent and is favoured by deprotonation of the amine - the degree of deprotonation increases as the pH level rises above 7.

42. However the skilled chemist would also have known that the monochloroacetic acid (MCA) will hydrolyse to glycolic acid as the pH level is increased above 7. Thus there is a balance or compromise which the skilled man would seek to achieve namely alkaline conditions but not too alkaline to destroy the MCA. (See Jurges XEX Day 1/74-76).

(1) UK Patent No. 1185111 (Morris)

43. The subject matter of Morris relates to the quaternisation reaction of alkylamines with MCA to form alkylbetaines in high yields.

44. The evidence was that the skilled person would consider such teaching to be relevant to the corresponding amidobetaines the subject of the pH patent (see Grigg 1 page 7; Jurges 1 paras 46-48).

45. Morris teaches one to maintain a pH within the range of 7-8 especially 7.5-8 throughout the reaction (p.2/29-32). A pH below 7 restricts the reaction whilst a pH higher than 8 tends to promote hydrolysis of the MCA. For maximum conversion an excess of MCA is required preferably a 20-25% molar excess depending on the temperature and time. Preferred temperatures over 70^oC, particularly over 80^oC are referred to - in particular boiling under reflux.

46. Examples 1 and 2 are conducted at a pH of 7.5 to 8 under reflux whilst Example 3 was only heated to 80^oC - all with excess MCA and conversions of amine to betaine of about 99% or better are achieved. At page 2/41 - 45 it is stated that a wider pH range of 6 to 9 can be tolerated for short periods at the expense of a slight loss of reaction of the tertiary amine - no doubt due to the increased hydrolysis of the MCA which can be rectified by adding more MCA.

47. It is contended by Goldschmidt that Morris teaches no more that what would be common general knowledge to the skilled man.

48. What Morris makes clear to one wishing to avoid residual levels of amine is to conduct the quaternisation reaction at an alkalinity level of preferably 7.5 to 8 at or just below the boiling point (80-100^oC) using a 20-25% molar excess of MCA. Such a mode of operation applied to the conversion of amidoamines to amidobetaines falls within the ambit of claims 1 and 2 of the pH patent - operating at a pH of 8 being within claim 2. There is no restriction in the pH patent as to the quantity of MCA that should be used or the level of glycolic acid that may result if the pH level exceeds 8.

49. Claims 1 and 2 of the pH patent are therefore obvious and lack inventive step over the Morris disclosure and/or the common general knowledge in the art.

14. (2) German Patent Application No. 2853171 (Phillips)

50. The subject matter of the Phillips disclosure is the preparation of a G phase comprising forming a betaine surfactant in the presence of another surfactant. Both alkyl and amidobetaines are disclosed. In the examples during the formation of the betaine pH levels of 7.5 - 8.0 and 8 - 8.5 respectively are maintained at a temperature of 65^oC. Such reaction conditions are peculiar to the formation of the G phase and the most that can be derived from such disclosure is that in the preparation of such systems, an alkaline pH of 7.5 - 8.5 was adopted and good conversions of amine to betaine were achieved.

51. Whilst such a teaching is consistent with both the Morris teaching and the common general knowledge I do not consider the Phillips disclosure on its own would invalidate the pH patent.

52. The Temperature patent claims a priority date of 26 February 1992 and relates to the preparation of amidobetaines by the quaternisation reaction of certain amidoamines with i.a. monochloroacetic acid (Step 2 supra).

53. It addresses the problem of organically bound chlorine in particular amidobetaines free of sodium mono and dichloroacetate (MCA and DCA). It is the presence of DCA as an impurity in the MCA starting material that the temperature patent primarily relates to.

54. Whilst use of higher pH will reduce the levels of MCA as taught by the earlier pH patent, levels of pHs higher than 10.5 not only give rise to problems of hydrolysis of the MCA starting material but also to the decomposition of the betaine product.

55. At page 2 line 30 reference is made to the prior art (DE-A3939264 (Uphues)) according to which the betaine is post-treated to react the DCA with ammonia, an amino acid or an oligopeptide.

56. Instead of this extra processing step, the Temperature patent merely requires the reaction to be carried out within a temperature range of 115^o to 180^oC.

57. At page 4 it is pointed out that such a temperature will necessitate that the reaction is carried out in a closed (pressurised) system such as an autoclave.

58. The lower limit of 115^o is stated to relate to the onset of the breakdown of DCA whilst the upper limit of 180^oC is determined by the onset of decomposition of the betaine product or the amidoamine reactant. At page 4 lines 18 to 22 it is stated that by means of the process it is possible to decrease the MCA and DCA levels in the betaine solution to below the detection limit of Ð 10 ppm.

59. Claim 1 of the Temperature patent (the only claim in issue) is to a process for the preparation of amidobetaines by quaternisation of the specified amidoamine with a w -haloalkylcarboxylic acid (or salt thereof) in aqueous solution characterised in that "the reaction is carried out within a temperature range from 115 to 180^oC until organically bound chlorine is no longer detectable, if appropriate the fatty amidodialkylamine being partially or completely quaternised in a first stage at 80 to 100^oC".

60. There is no dispute that the expression "no longer detectable" relates to levels of MCA and DCA respectively of less than 10 ppm as stated at page 4.

61. It is to be noted that the claim is not limited to any pH range and indeed in Example 2.3 no sodium hydroxide is added to maintain the reaction medium alkaline.

62. The only evidence Goldschmidt put forward as to the infringement of the Temperature patent is in respect of Sample X which is a sample of C45 supplied to Surfachem by EOC having a certificate of analysis dated 25 July 1996.

63. The production temperature recorded for runs relating to such sample are set out in PD8 as follows:

64. The only production run which may fall within claim 1 of the Temperature patent is Run 284 if the recorded temperature of 120^oC (or a temperature above 115) was maintained until the DCA/MCA levels were no longer detectable. From the records available Run 284 was a 20 tonne batch which was treated at 120^oC for at least two hours (see Delahaye Day 4 524⁶-525²³). Samples D,E and F which Goldschmidt acquired in Germany and France at about the same period (June/July 1996) were analysed by Goldschmidt and found to have DCA levels below 10 ppm. Either such low levels are the result of EOC's use of high temperature (120^oC for at least two hours) or the DCA content of the MCA reactant was such that the level of DCA fell below 10 ppm in the final betaine product. I consider the evidence of Mr. Delahaye that Hoechst may have been supplying MCA having such low levels of DCA to be speculative (Day 4/519⁷-521¹⁷) and that the likely reason for such low levels of DCA in 1996 was that at least for certain runs (e.g. Run 284) a temperature of 120^oC was used in the EOC autoclaves. EOC have refused to provide discovery relating to Samples D, E or F as to the temperatures that were recorded for such samples. EOC Product-Method Sheets for the preparation of C45 in 1995 and 1996 specify as part of the process that the reactor is to be set to 1.5 bar nitrogen and continue heating to 120^oC. They further specify that "Due to heating the pressure will increase to 2 bar. After 2 hours at 120^oC send sample to lab". I consider that on the balance of probabilities that Run 284 would have had a level of organically bound chlorine which was below 10ppm (no longer detectable) and that the cause of such low levels of MCA/DCA would have been due to its treatment at 120^oC for at least two hours.

65. It follows that in so far as Sample X contained a proportion of amidobetaine prepared from Run 284, such product fell within claim 1 of the Temperature patent.

66. I consider runs at temperatures below 115^oC do not fall within such claim as it is made clear that the temperature window of 115 to 180^oC is indeed essential to the working of the invention.

67. By 1992 the presence of DCA impurity in the MCA used to react with the amidoamine to produce the amidobetaine was common knowledge. In 1991/1992 the personal care industry had expressed its concern as to the carcinogencity of DCA (see Jurges Day 1/65-66, Day 2/158). Hence by 1992 a skilled person seeking to manufacture such amidobetaines would be concerned to minimise not only the level of MCA in the final product (a matter which the pH patent addresses) but also the more intractable problem of his DCA levels - given that the supplier of MCA was unable to do so or only able to do so at an unacceptably high price.

68. In their closing submissions EOC limited their invalidity attack to one citation, namely an article by F. Kunze in Z. Phys. Chem 188, 99-108 (1941) entitled "On the hydrolysis of dichloroacetic acid". It was common ground that the notional skilled person would do a literature search and consider the Kunze article.

69. At page 9 of the translation (Table 1) rates of hydrolysis of DCA at temperatures of 100^o, 114^o and 123.5^oC are tabulated. From this Table it is clear that the rate of hydrolysis will increase significantly as the temperature is raised from 100^o to 114 and even more so from 100 to 123.5^oC. In such Experiments no sodium hydroxide is added, DCA is merely heated in water.

70. Goldschmidt contend that the rates disclosed are exceedingly slow. In order to remove DCA down to what was considered to be acceptable levels (10 to 20 ppm) unacceptably long treatment times would be required (see Dr. Jurges First Report pages 87 to 90).

71. However at pages 5 and 6 of Kunze, data is provided as to the hydrolysis of DCA in the presence of sodium hydroxide at 123.5^oC and is compared with the case without additive. It can be seen that the level of hydrochloric acid (produced when DCA is hydrolysed) has increased by a factor of 4 over the space of 4 hours. In other words the Kunze article teaches that whilst an increase in temperature alone will not be sufficient to reduce the DCA impurity to an acceptable level - such an increase in temperature to say 123.5^oC coupled with the addition of sodium hydroxide would be of the order of four times more effective even over a period as short as 4 hours (cf. the Examples in the Temperature patent vary from 4 to 8 hours).

72. I consider the Kunze teaching at pages 5 and 6 would provide a solution to the person presented with the problem of removing DCA - at least it teaches the reader this is one way of significantly lowering the DCA levels.

73. Goldschmidt contend that if one were to follow the Kunze teaching one would require a pressure vessel or autoclave and that in this art the industry generally did not use such vessels. In particular if one did one would face difficulties of adding the sodium hydroxide to keep the pH in the desired range. I find this contention difficult to understand in that EOC do use such vessels and it has not been suggested there is any particular difficulties in such use.

74. In any event the fact that there may be commercial reasons not to use pressure vessels unless one has to, is no answer to it being technically obvious to do so - whether in the laboratory or on an industrial scale - see Grigg Day 3/416¹⁰ - 418^19.

75. Some suggestion was made that at the higher temperature and at higher pH there would not only be a tendency to hydrolyse the DCA but also the betaine. In practice no significant hydrolysis of the amide bond will occur below a pH of 10 - see Grigg Day 3/406²⁶- 408¹³. Indeed the pH patent teaches a pH range of 8-10 is required for removal of MCA. I do not consider knowledge of the possibility of hydrolysis of the betaine would deter one skilled in the art from at least trying to hydrolyse DCA as taught by Kunze under laboratory conditions.

76. I consider that claim 1 of the Temperature patent is an obvious solution to the removal of the unwanted DCA in view of the teaching in the Kunze article when the quarternisation reaction is carried at a pH of eg. 7 to 9 - there is no pH limitation claimed.

77. Claim 1 of the Temperature patent is obvious. No separate inventive step is alleged in respect of any of the appendant claims which I hold are also obvious over Kunze.

78. The Concentration patent claims a priority date of 9 March '92. It is based on the empirical discovery that the addition of small quantities of fatty acid (1-3% by weight) will allow a significant increase in the solids content of an aqueous solution of an amidobetaine whilst still remaining liquid.

79. Solids content is defined as the amount by weight of solids determined by evaporating a sample for two hours at 105^oC. The patent explains that known aqueous solutions of betaines are only liquid below a certain concentration of solids depending on the fatty acid used in its preparation. Typically a betaine derived from coconut fatty acid solidifies at a solids content of about 40% by weight and consequently commercially available aqueous solutions of such betaines as at the priority date had a solids content of about 35% by weight.

80. The reason for having a higher solids content is not simply one of economics (less water to transport) but it is stated that to avoid the need for preservatives against bacterial decomposition a solids content from 40% by weight is required (see page 5 lines 17 to 21, page 10 lines 2 to 3 and Example C).

81. Reference is made to prior art attempts to increase the solids content which include German Patent Specification 3726322 (which corresponds to US Patent No. 4,861,517 (Bade) cited by EOC against the Concentration patent). This involves the addition of a mineral acid such as hydrochloric acid to reduce the pH of the betaine solution to a pH of from 1 to 4.5.

82. At page 5 lines 22 to 26 it is stated that the achievement of increasing the solids content of an aqueous liquid betaine solution to at least 40% by weight with a pH of from 5 to 8 without additional processing steps is "carried out by the setting of a certain content of free fatty acid and, if appropriate, small amounts of glycerol in the solution of the betaine".

83. At page 9 lines 17 to 20 the patent states that "If appropriate the pH is adjusted to 5 to 8 using a suitable acid after completion of the reaction before cooling to room temperature. The pH range from 5 to 7, in particular from 5.5 to 6.5. is preferred".

84. Comparative Examples (A1 and A2) having fatty acid contents both below and above the claimed range result in non fluid products. Similarly comparative Example A4 results in a solid product even though it has a fatty acid content within the claimed range due to the fact that there was no pH adjustment after completion of the reaction by the addition of a mineral acid so that the pH remained at 8.6.

Claim 1 is to an aqueous liquid solution of an amidobetaine of the general formula

where R= alkyl radical of preferably hardened coconut fatty acid or a fatty acid mixture which on average corresponds to coconut fatty acid, the solution having:

# a solids content of at least 40% by weight

# a pH of 5 to 8 and

# an aminoamide content of £ 1 % by weight

characterised in that

# it contains 1 to 3% by weight (based on solution) of one or more saturated fatty acids having an average 8 to 18 carbon atoms or one or more unsaturated fatty acids having an average 8 to 24 carbon atoms and

# 0-4% by weight (based on solution) of glycerol.

Claim 2 is limited to a range of 1.5 to 3% by weight of fatty acid.

Claim 6 is to a process for the preparation of a betaine solution according to one or more of the preceding claims which is characterised in that a fatty acid aminoamide which contains the desired amount of free fatty acid is used for the quaternisation reaction, or the desired amount of fatty acid is added to the reaction mixture before or during the quaternisation reaction and - if appropriate glycerol is added to the reaction mixture.

85. The main dispute between the parties is as to the proper interpretation to be given to the phrase "the solution having … a pH of 5 to 8".

86. Do such words limit the scope of Claim 1 (and therefore all appendant claims) to a window of 5.0 to 8.0 (as EOC contend) or do such words mean 5 to 8 rounded to the one significant number.

87. Mr. Fysh Q.C. has referred me to the Protocol to Article 69 EPC that the claims are to be interpreted so as to combine a fair protection for the patentee with a reasonable degree of certainty for third parties. In the context of numerical ranges I was also referred to Auchincloss v. Agricultural & Veterinary Supplies Ltd (1997) RPC 649 at 664 and 689 where Peter Prescott Q.C. (sitting as a Deputy judge) distinguishes a claim in which there are numerical limitations from a claim where a descriptive phrase is used.

88. Mr. Prescott there considered that the patentee by chosing a numerical limitation had circumscribed his monopoly to such numerical values. Thus far and no further. To hold otherwise would be to deny a reasonable degree of certainty for third parties.

89. I consider that the above observations have to be viewed in their context. In the Auchincloss case the claim was to a biocidal composition comprising a formulation of a number of ingredients in varying amounts. It was a recipe in which each of the ingredients were stated to be used in amounts within the specified ranges.

In Lubrizol Corp. v. Esso Petroleum Co. Ltd (1998) RPC 727 at 748 the Court of Appeal (Aldous L.J) held that the notional skilled man would read the claim and the specification using conventions adopted by scientists, one of which was that numbers are given to the number of figures that are significant. In that case the claim to "at least 1.3 succinic groups" was held to include 1.28 or 1.29. In other words it was construed not as a claim to at least 1.30 succinic groups but to 1.3 rounded.

90. In the present case the pH range is 5 to 8 i.e. to one significant figure and not 5.0 to 8.0.

91. The evidence is that pH is generally measured by a pH meter and in an industrial plant to one decimal point. The pH values for each of the Examples in the patent are also recorded to one decimal point. This is to be contrasted with the claimed pH range of from 5 to 8.

pH measurement of a surfactant solution having a high solids content will be subject to variability. The European Standard (X9) refers to a repeatability of 0.2 of a pH unit and a reproducibility of 0.3. Dr. Jurges considered there to be an area of uncertainty of 0.3 to 0.5 units (Day 1/115, Day 2/185^10-20). Whilst Mr. Bohmer (Goldschmidt's analytical chemist) put the variability between any two measurements of the same sample as 0.2 or 0.3 units (Day 2/237-238). Prof. Lilley who had no experience in measuring the pHs of surfactant solutions having a high solids content was surprised to see a 0.3 difference in measurements.

92. I consider that one skilled in the art when viewing a pH range of 5 to 8 would not have read such figures as being 5.0 to 8.0 but would have understood them to be to one significant figure only. Such an interpretation is consistent with the reference in the Concentration patent to German Patent Specification 3726322 (Bade) and the treatment of a betaine solution with a mineral acid to reduce the pH to a value of from 1 to 4.5 which is stated to give rise to corrosion problems during storage and transport (See Dr. Jurges' first report para 112). It is also consistent with comparative Example A4 having a pH of 8.6. In other words when construed purposively, the lower limit pH of 5 is to avoid corrosion problems caused by a pH of below 4.5 and the upper pH limit of 8 is to avoid solidification above a pH of 8.5.

93. Both C45 and C47 are alleged to infringe the Concentration patent. Goldschmidt accept protection for the Concentration patent only commences on the date of its grant namely from the 17 April 1996 as no translation of the claims of the application were filed prior to grant (see section 78(1)-(3), (7) and section 69).

94. EOC use two different types of amidoamine feedstock to produce their C45 and C47 betaine. K2498 is produced directly from coconut oil whereas K2499 is produced from a mixture of fatty acid methyl esters. Both feedstocks contain certain residual amounts of fatty acid. K2499 will also contain unreacted methyl esters which will hydrolyse during the quaternisation reaction to produce further fatty acid in the final betaine product.

95. To determine the total amount of free fatty acid in the betaine solution it is necessary to take account of the fatty acids in the amidoamine feedstocks used as well as any added fatty acid.

96. EOC have over the years changed their methods of production of C45 and C47 in the following material ways:

1. From November 1995 to March 1997; in the production of C45, 0.5% by weight of fatty acid was added. The pH specification was 4.5-5.5 until June 1996.

2. In June 1996 in response to a letter before action from Goldschmidt the pH specification for C45 was changed to 4.5 - 5.0.

3. In May 1997 on the commencement of the production of C47 the fatty acid added was reduced to 0.3% by weight.

4. From the lst October 1997, 2% citric acid was added.

5. Since August 1998 (i.e. post writ) no fatty acid has been added; instead a mixture of K2498 and K2499 feedstocks has been used. All fatty acid present in the betaine solution is derived from such feedstocks.

97. Set out hereunder are the C45, C47 betaine products which have been imported to the UK by EOC and which are alleged to infringe the Concentration patent together with the relevant data (taken from EOC's Table marked D2 and Goldschmidt's summary appended to their closing submissions).

98. So far as the free fatty acid levels are concerned, the EOC Certificates of Analysis only record the pH, solids content and levels of MCA. To determine their fatty and levels Goldschmidt conducted experiments on a number of samples of EOC products including samples B and C.

99. Some of the C47 samples tested by Goldschmidt (not the subject of the Notice of Experiments) had fatty acid levels below 1%. However there is no available data as to the precise recipe of such samples viz. the amidoamine feedstocks used.

100. Mr. Delahaye (para 37 of his second Report) states that until mid 1997 EOC did not have a method for testing for fatty acid content. Thereafter sample runs were tested from time to time.

101. Mr. Bohmer has performed a calculation (X/2) to determine the amount of free fatty acid which the amidoamine feedstocks K2498 and K2499 will contribute to the final betaine product. Thus for any given recipe one can determine the fatty acid content of those products supplied by EOC to Surfachem for which Goldschmidt were unable to obtain samples. Applying such calculation, C45 batches such as samples X and Y had a fatty acid level of about 1.3% (which compares favourably with measured samples of C45 e.g. Samples D and E of the same date). Whereas the C47 batches where the level of added fatty acid had been reduced from 0.5% to 0.3% by weight had a fatty acid level of about 1.1.% which is consistent with the C45 levels given that 0.2% less fatty acid was added.

102. As regards the position since August 1998 (post writ) so far as it is relevant to this case Goldschmidt were provided with a sample (sample Z) of a production run demonstrated to them by EOC which has a fatty acid level of 1.03% (Goldschmidt measurement) and 0.9% (EOC measurement). Much criticism was directed to the EOC measurement (see Bohmer Report paras 110-129) in particular as to the absence of any internal standard.

103. No one from EOC was called to support their measurement.

104. Mr. Bohmer believes his measurement to be accurate to + 0.5% relative (Bohmer I para 141; Day 2/249) i.e. a measured fatty acid content of 1.03% + .05 and hence if and in so far as sample Z is said to be typical of EOCs production run as at August 1998 (using a mix of amidoamine feedstocks) it will have a free fatty acid content of about 1%.

105. EOC contend that the Bohmer calculations to prove infringement cannot be relied upon although they do not dispute that all of the C45 supplied to Surfachem in the U.K. had a fatty acid content greater than 1% (see Delahaye Day 4/480 5-12). In particular EOC say that the fatty acid values determined for K2498 and K2499 of 4% and 2.8% respectively were at best approximations as to which Mr. Bohmer was unable to say what degree of certainty could be attached to them (see Day 2/241-242). Thus (it is said) where the fatty acid level approaches 1%, such calculations for determining the fatty acid levels of the C47 products cannot be relied upon.

106. No relevant samples for C47 products supplied to Surfachem have been made available to Goldschmidt and it has not been possible to correlate any recipe with a measured fatty acid content. EOC could have refuted the Bohmer calculations by measuring the actual levels of free fatty acid in K2498 and K2499 in properly conducted experiments. I consider the Bohmer calculations are the best evidence available and as already stated the values for C47 for the COM 1 and COM 2 samples are consistent with the measured values obtained for C45 given that 0.2% less fatty acid was being used.

107. EOC further contend that it is the addition of citric acid which fluidises the betaine solution rather than the free fatty acid content.

108. Mr. Delahaye (first Report paras 27 to 28) explains that when making C45, citric acid was used to adjust the pH to bring it into the range of 4.5 to 5.5 (and from June 1996 a range of 4.5 to 5.0). However when making C47 (which had 0.2% less added fatty acid) it was found that some of the batches failed the cold test (in accordance with which the product had to remain fluid at 10^oC). This was rectified by adding more citric acid and if the pH went below 4.5 adding sodium hydroxide to bring the pH back up again. This procedure was followed from May to September 1997. However since October 1997 EOC have routinely added a fixed amount of excess citric acid (about 2% by weight) which will lower the pH below 4.5 which is then brought back to a pH within the 4.5 to 5.0 range by the addition of sodium hydroxide (referred to as citric acid overshoot).

109. The fact that such levels of citric acid may have a beneficial fluidising effect will not avoid infringement of the Concentration patent if free fatty acid is also present in the claimed range of 1 to 3%.

110. Finally so far as the pH values are concerned for the samples alleged to infringe which range from 4.6 to 4.9 (excluding sample Y which has a value of 5.2), all such values when rounded to one significant number result in a pH of 5 and therefore fall within claim 1 on the construction I have found to be the proper one.

111. However even assuming that the proper construction of a pH of 5 to 8 is to be interpreted as meaning to one decimal point (e.g. 4.95 is rounded to 5.0) and applying the "Improver Questions", I do not consider the pH values of 4.6 to 4.9 to be excluded from the scope of such a claim.

112. There is no material difference in the fluidity of a betaine solution whether the pH is 5.0 or 4.9 or indeed 4.6 and I consider a skilled person in 1993 would have understood that to be so. The Concentration patent points out that a pH of 1 to 4.5 is too corrosive. Dr. Jurges (First Report para 112) considered that a solution was only viewed in the industry as being strongly acidic below a pH of 4 to 4.5. Furthermore in spite of the change of pH in June 1996 by EOC (from a range of 4.5 to 5.5 to 4.5 to 5.0) Surfachem continued to represent that the product they were supplying to their customers typically had a pH of 5 with a specification of from 4.5 to 5.5 (See Bundle G/ Tab 6 p. G.32).

113. As to whether by claiming the range of 5 to 8 the patentee intended to exclude pH values below 5, I do not believe that to be so - had they so wished to exclude values such as 4.6 to 4.9, the claim would have been more precise and claimed a range of 5.0 to 8.0.

114. For the reasons stated above , I consider Samples X , Y, B, C and Com 1 and 2 fall within claims 1 and 6 of the Concentration patent.

115. Surfachem seeks a declaration that a product having a free fatty acid of less than 0.95% by weight would not infringe the Concentration patent (Defence and Counterclaim para 12).

116. This declaration does not comply with section 71 of the Patents Act, 1977 which requires full particulars in writing of the act or proposed act.

117. The letter from Walker Morris dated 11th June 1999 (Bundle B Tab 8 page 16) merely describes a hypothetical product. There is no evidence that Surfachem/EOC proposes to manufacture or sell betaine with such a level of free fatty acid and in particular what recipes are used to produce such a product. Indeed no recipe has been provided for Sample Z which I have found to have a free fatty acid content of about 1% (1.03 + .05%).

118. There is simply not enough information provided by EOC in order for the court to consider any actual or proposed product in terms of its measured or calculated free fatty acid content.

119. A declaration of non infringement in respect of a product as to which there is no evidence as to whether it can be made or consistently made would be purely theoretical and one the Court should not entertain.

120. The Concentration patent is alleged to be obvious over US Patent No. 4861517 (Bade) and the common general knowledge in the surfactant field.

121. It is also alleged to be insufficient. Claims 1, 2 and 6 are alleged to be independently valid.

122. Prof. Lilley sets out the general knowledge that he perceives a chemist working with surfactants would have had in 1992 (see section 3 and Appendix 2 of his Report). In summary he states

• surfactant molecules when in relatively dilute solutions in water aggregate to form spherical micelles.
• at higher concentrations, rod-like aggregates and sheet-like lamellae can form.
• the viscous behaviour of surfactant systems depends on whether spherical micelles, rods or lamellae are present.
• the physical properties such as viscosity of systems containing surfactants depends on temperature and pH.

123. EOC's main attack is that it was known that the micellar solutions can be stabilised and therefore remain at higher concentrations by the incorporation of another suitable surfactant to form mixed micelles. However as acknowledged by Prof. Lilley (Day 3/356^2-19) surfactant systems are extremely complex and what may apply to one type of detergent species will not necessarily apply to another. There are anionic, zwitterionic (such as betaines), cationic and non-ionic species which may be admixed with the primary surfactant.

124. EOC contend that a fatty acid at a pH above 5 is similar to a soap which is an anionic surfactant and therefore would be obvious to try.

125. Fatty acids were known as thickeners for soaps and surfactant solutions (see Dr. Jurges first Report para 118 and Prof. Lilley (Day 3/352¹⁷-355²⁴).

126. Thus adding a fatty acid might lead to an increase in viscosity due to its likely behaviour as a small headed non-ionic surfactant (see Prof. Lilley Day 3/351⁶-352¹⁵) or when dissociated (i.e. at higher pH levels) at best it would be uncertain whether the charged species would compensate for the small head (Day 3/340¹¹-341³).

127. I do not consider the concept of utilising 1-3% of a free fatty acid to enable a higher solids content solution of a betaine surfactant to be obvious in the light of the common general knowledge in the art as of 1992. Indeed Dr. Jurges expressed surprise when it was first heard of at Hoechst (see Dr. Jurges First Report para 118). Given the benefits of a higher solids content, if it were obvious it is not clear why no one had thought of doing it before in spite of various attempts to do so (see Prof. Lilley Day 3/362³-364¹⁶).

128. Bade is referred to at page 4 of the Concentration patent. It has as its object the provision of a "simple and economic method of preparing concentrated, flowable aqueous betaine solutions" having a solids content of 43 to about 65% by weight. In its simplest form mineral acid is added to the solution to bring the pH down to between 1 to 4.5.

129. At Col 4 lines 43 to 51 a preferred embodiment is referred to in which 1 to 5% by weight of a water soluble non ionic surfactant is added before the pH is adjusted to a value of 1 to 4.5. The betaine solution so produced may then be adjusted to a pH of ³ 5 to 9 by addition of alkaline solution.

130. A fatty acid is not really a surfactant - it does not aggregate or reduce surface tension significantly. It is certainly not a water soluble nonionic surfactant - it is either ionic in its dissociated form at the higher pH levels or it is not water soluble in its undissociated from (see Dr. Jurges First Report para 116). There is no suggestion in Bade to use an anionic or cationic surfactant. I do not believe the evidence begins to establish it would be obvious to replace the nonionic surfactants disclosed with a fatty acid. As Prof. Lilley stated (Day 3/356^2-19)

131. In paragraph 2 of the Re-Re-Amended Particulars of Objections there are two matters which are alleged to render the Concentration patent insufficient, namely

(1) how to measure fatty acid content

(2) the claims include products that the patent does not teach how to make (the Biogen point)

132. As regards the first matter, I did not understand this to be pursued by EOC. Indeed Mr. Bohmer's evidence as to the various ways that the fatty acid content could be determined was unchallenged.

133. As regards the Biogen point, it is contended that in so far as the C47 products are rendered fluid by virtue of the added citric acid they owe nothing to the technical contribution of the concentration patent, namely fluidisation by virtue of having 1 to 3% by weight of free fatty acid.

134. EOC conducted experiments to establish that citric acid has a fluidising effect. Goldschmidt conducted Experiments in Reply to show that fatty acid also has a fluidising effect - indeed Prof. Lilley (Day 3/366^21-24) accepted as much:

"Yes, I would imagine that this (i.e. fatty acid) is always making a contribution …"

135. Because EOC's C45 and C47 products contain levels of free fatty acid between 1% to 3% by weight, I have no doubt they are making use of the technical contribution the subject of the Concentration patent - even though such products may have enhanced properties due to the addition of citric acid.

136. I do not consider the Biogen point is relevant to this case. The patent is not insufficient.

137. In the light of my finding that the Concentration patent is both valid and as from the 17th April 1996 infringed by EOC and Surfachem both as regards the C45 and the C47 products it is accepted that Goldschmidt will have a defence under section 70(2) of the Patents Act 1977 in respect of the alleged threats made by Goldschmidt U.K. against EOC and Surfachem (even though such threats may also relate to other Goldschmidt patents in respect of which they could not be justified).

138. By their closing submissions, EOC had refined their threats allegations mainly to certain letters sent by a Mr. Davies of Goldschmidt UK to its customers.

139. The commercial background to such letters was that by mid 1996 Goldschmidt were concerned to promote their high solids betaine product referred to as Tego Betaine F50 but were aware of EOC's C45 product (known also as DEV 112) which Mr. Davis had been told infringed the Concentration patent.

(1) The first circular letter (B Tab 8 p. 11-12) was sent in August 1996 by Mr. Davies to thirteen of Goldschmidt's UK customers or potential customers who were first spoken to before the letter was sent. The letter draws the attention of the reader to the Concentration patent which is said to relate to the F50 product and the fact that EOC Belgium were offering in the UK via their agent Surfachem a product covered by the patent. The letter goes on to state that EOC had been notified and it concludes by stating that as a highly valued customer, "we are concerned that a situation has arisen whereby you may have been offered this product. We feel it is only fair to notify you of this situation. If you have any concerns about sourcing for Cocoamidopropyl Betains we would be glad to discuss this subject with you".

140. The underlying message in this circular was that Goldschmidt would be suing EOC Belgium and Surfachem under the U.K. Concentration patent in respect of the supply of EOC's C45 product (see Davies Day 2 284-288). The Writ in these proceedings was not issued until 23rd January 1998 - although saisie proceedings in Belgium were commenced on 12 February 1997. I have no doubt that this letter constitutes a threat made to Goldschmidt's customers to sue EOC Belgium and Surfachem under the Concentration patent in accordance with section 70(1) of the Patents Act 1977 and would have been so understood by the recipient.

(2) Mr. Davies visited Topaz as Topaz had taken the EOC/Surfachem B45 product. Mr. Davies agreed that he told Topaz or that Topaz would have understood that proceedings were likely to be brought in the U.K. against Surfachem and EOC Belgium in respect of the B45 product (see Davies Day 2/289-293).

141. I consider this is a clear further threat made to Topaz to sue EOC Belgium and Surfachem in accordance with section 70(1).

(3) In late 1997 Goldschmidt U.K. sent out a second circular letter (B Tab 8 p. 10) to a some 32 customers or potential customers. Some of the recipients were the same as those who received the first circular. Again all recipients were contacted in advance.

142. The letter states that Goldschmidt A.G. has taken legal steps to enforce its patent rights and that Surfachem, EOC Belgium's U.K. distributor has been informed of the situation.

143. By this time the Belgium proceedings had been commenced although to what extent the U.K. recipients of this letter were aware of that fact is not clear. The reference to Surfachem having been notified of the situation seems to imply that they are being targeted in the U.K. if they continue with their activities and therefore to that extent is a further threat against them.

144. It follows that subject to the fact that Goldschmidt have established such threats were justifiable, they would have been unlawful.

145. Mr. Floyd Q.C. for the Claimants pointed out that so far as expertise is concerned, Prof. Grigg for EOC had no experience in the manufacture or formulation of surfactants (See Day 3/398^14-19) and Prof. Lilley had very little experience of surfactants (See Day 3/321³-323²⁵). It was said that as such neither expert could be considered to be the relevant skilled addressee. Whereas Dr. Jurges for Goldschmidt (so it was contended) was such a person with hands on experience.

146. This submission was made with particular regard to the obviousness questions and the differing approach someone outside the industry may take from someone inside the industry.

147. On the other hand Mr. Fysh Q.C. stated he made no attack on the credibility of the Claimants witnesses although he submitted that even allowing for linguistic difficulties on occasions some of Dr. Jurges answers were both prolix and obscure. I have borne in mind both parties' observations when addressing the issues in this case.

CONCLUSION

1. European Patent No. 0560114 (the Concentration patent) is valid and has been infringed by the importation, sale and supply of EOC's C45 and C47 products in the U.K.

2. European Patent No. 0020907 (the pH patent) is invalid.

3. European Patent No. 0 557 833 (the Temperature patent) is invalid.

4. The declaration of non-infringement is refused

5. The threats made by Goldschmidt UK concerning EOC Belgium and Surfachem are justified.