Potassium (K) is a primary constituent of planetary crusts and studies of its stable isotope ratio (41 K/ 39 K) have been used recently to provide new insights into planetary accretion and differentiation. Experimental work shows that, as a moderately volatile element, K isotopes fractionate substantially during evaporation-condensation processes (e.g., Yu et al., 2003). Measurements of differentiated lunar and terrestrial lavas indicated that the Moon is enriched in heavy K isotopes by ∼0.4‰ relative to the Earth, which has been interpreted as evaporative loss of light K during the Moon-forming giant impact (Wang & Jacobsen, 2016a). Such inter-planetary comparison is based on a key assumption that negligible K isotope fractionation occurs during igneous differentiation; therefore, crustal rocks are assumed to be representative of their planetary bodies. This assumption is supported by less equilibrium isotope fractionation as temperature increases (∼1/T 2 , Bigeleisen & Mayer, 1947; Urey, 1947). Also, K has a strong tendency to partition into melts during mantle melting and magmatic differentiation. Hence crystal-melt fractionation is not expected to significantly influence the K budget and isotopic composition of evolving magmas. However, despite these theoretical considerations, there is sparse confirmation in natural samples. Potassium occurs in several major silicate groups that are commonly involved in magmatic differentiation. Plagioclase crystallization is widely associated with basaltic differentiation while crystallization of K-rich hornblende, biotite, and K-feldspar occurs in more evolved magmas. The sparse δ 41 K data reported for these minerals span a range of 0.6‰, with biotite being isotopically lighter than typical igneous rocks and feldspar being mostly heavier except a few outliers (Chen, Tian, et al., 2019; Morgan et al., 2018). Their contrasting preferences for K isotopes could reflect differing K coordination environments between crystallized minerals and melts, or among various K-bearing minerals. Theoretical calculations also suggest that the K-O bond strength of a mineral varies with its chemical composition, for example, feldspar with lower K/(K + Na)