On injectivity of maps between Grothendieck groups induced by completion (Q841527)

Let \((R,m)\) be a local ring, \({\hat R}\) the \(m\)-adic completion of \(R\) and \(f:G(R)\rightarrow G({\hat R})\) the map induced by base change between the Grothendieck groups. \textit{Y. Kamoi} and \textit{K. Kurano} [J. Algebra 254, No. 1, 21--43 (2002; Zbl 1095.13523)] showed that \(f\) is injective if either \(R\) is Henselian, or \(R\) is an isolated singularity, or \(R\) is the localization of the irrelevant ideal of a positive graded ring over a field. Hochster gave an example when \(f\) is not injective but \(R\) is not normal. Here it gives an example when \(R\) is normal but \(f\) is still not injective.