Numerical limitation inventions are inventions whose point of novelty or distinction from the prior art lies in a limited numerical range of values that defines a previously disclosed genus or element (i.e., in this type of inventions, one or more of an invention’s technical elements are defined by a numerical range).  In its decision in Case 2019Hu12094 dated January 13, 2022 (hereby referred to as the “Decision”), the Supreme Court held that the assessment of the inventiveness of a numerical limitation invention should be based not only on descriptions disclosed in the prior art, but also on matters reasonably recognizable by a person skilled in the art based on the entire prior art.

Previous Supreme Court decisions have ruled that a numerical limitation invention should exhibit a qualitatively different or quantitatively remarkable effect over the prior art to be valid.  Generally, if the objective of the numerical limitation invention is the same as that of the prior art, and the numerical limitation itself is the only difference from the prior art, the Supreme Court used to find the invention to be inventive and thus valid over the prior art only if the adoption of the numerical limitation resulted in such a different or remarkable effect.  As a result, courts have tended to review the inventiveness of numerical limitation inventions by simply observing whether there was any significant difference in effect within the limited numerical range, regardless of the context of the element in the prior art that was limited.  The recent Supreme Court Decision, however, appears to present a significantly different standard for reviewing inventiveness.

The patent-applied invention at issue relates to a method of treatment of a running ferrous alloy sheet to improve the adhesion of coatings on the sheet.  The invention involved two numerical limitations concerning the viscosity of the “molten oxide bath” used to remove the oxide on the surface of the sheet to improve coating adhesion, and the content ratio of Li₂O in the “molten oxide bath.”

Numerical Range of “Melting Oxide Bath” Viscosity
 

For the limitation on viscosity, the prior art disclosed a molten salt bath with a viscosity up to 100 poises (without an explicitly stated lower bound), whereas the invention claimed a viscosity range of 0.003-3 poises.  Considering the above, the Intellectual Property High Court (the “IP High Court”) held that the numerical range was not inventive, explaining that the claimed feature had no technical significance since the two ranges were overlapping and as there was no data disclosed to show the amount of coating adhesion within the claimed numerical range.  However, the Supreme Court held that based on the overall description of the prior art, one of ordinary skill can reasonably figure out that the minimum viscosity for forming the coagulation coating should be the lower limit of its viscosity range.  The Supreme Court then held that, as lowering the viscosity in the prior art to the claimed numerical range would not have allowed a coagulation coating to form, one of ordinary skill could not have easily derived the invention.  Therefore, it ruled that, in view of the overall disclosure of the prior art, the claimed numerical range was inventive despite the overlap in the claimed numerical range and the range disclosed in the prior art.

Content Ratio of Li₂O
 

Similarly, regarding the content ratio of Li₂O, while the range of Li₂O (i.e., 10wt% ≤ Li₂O ≤ 45wt%) stated in the claimed invention did not actually overlap with the range disclosed in the prior art (i.e., not exceeding 6wt%), the IP High Court held that the claimed numerical range had no technical significance and could not be deemed inventive as (i) the invention and the prior art had the same purpose of reducing the amount of oxide adhered to the sheet’s surface by reducing the viscosity of the bath; (ii) the viscosity ranges of the two inventions’ salt baths overlapped; and (iii) there was no data to support any significant difference in effects within the claimed range.

However, based on the premise that the prior art’s disclosure that “Li₂O content exceeding 6.0 wt% deteriorates peelability and thus should be avoided” can be viewed as a negative instruction on the addition of Li₂O in an amount exceeding 6.0 wt%, the Supreme Court held that “it is difficult for a person skilled in the art to change the composition ratio of Li₂O in the prior art to 10 wt% ≤ Li₂O ≤ 45 wt% without ex post facto consideration of the above negative instruction or implication.”

This decision is significant in that it is different from the practice that has considered substantiating the remarkable effects of numerical limitation as the main standard in assessing the inventiveness of numerical limitation inventions that have the same objectives and effects as the prior art.

The key takeaway of the decision is that the Supreme Court has now ruled that assessments of the inventiveness of a numerical limitation invention must be based on whether an ordinary person skilled in the prior art could have easily derived the numerical limitation invention from the overall description of the prior art without hindsight bias, rather than being simply dependent on the “critical significance” of a numerical range (i.e., whether there is a significant difference in effects within compared to outside the claimed numerical range).  This ruling marks a clear shift from previous jurisprudence regarding numerical limitation inventions in Korea.  Therefore, in cases where the inventiveness of an numerical limitation invention is disputed in a patent application or invalidation case in the future, it will be important to accurately identify what factors a person skilled in the art would recognize from the entire disclosure of the prior art.

[Korean Version]