Substituting a native ion in the crystals with a foreign ion that has a difference in valence, termed as aliovalent doping, has been widely attempted to upgrade solid-state ionic conductors for various charge carriers including O2−, H+, Li+, Na+, and so forth. 1− 4 The doping aids to promote the high-conductive framework and dredge the tunnel for fast ion transport. The garnet-type Li7La3Zr2O12 (LLZO) is a fast Li+ solid conductor, which received vast attention as an electrolyte candidate for allsolid-state lithium ion batteries, showing great potential to offer high energy density and minimize battery safety concerns to meet extensive applications in large energy storage systems such as electric vehicles and aerospace. 5− 8 In the Li-stuffed garnet framework of LLZO, the 3D pathway formed through the incompletely occupied tetrahedral sites bridged by a single octahedron enables the superior Li+ conductivity. 9, 10 For the purpose of optimal performance, many efforts of aliovalent doping have been made throughout metal elements (Al3+, Ta5+) and metalloid elements (Ga3+, Te6+) in the periodic table with various valences 11− 14 to stabilize the high-conductive phase and increase the Li vacancy concentration. 7, 10, 15 However, the governing mechanism of the high conductivity through aliovalent doping is still not fully understood. Doping does not result in a much different garnet framework of highconductive cubic phase from that of the low-conductive tetragonal phase. 16, 17 The aliovalent doping does not tremendously change the Li vacancy concentration, either, because of the “sufficient” vacancies (∼ 16 vacancies distributed in the 3D-connecting 24 tetrahedral sites plus 48 octahedral sites) preexisting in both cubic and tetragonal phases although with different arrangements. 16, 17 The slight structure tuning above is not fully responsible, however, for ionic conductivity varying by orders of magnitude 16 among those “similar” garnets. It is noted that the vacancies in tetragonal LLZO reside in the tetrahedral sites but hardly help the conduction. Therefore, rather than the global concentration, the vacancy density in the right site, the bridging octahedral site, is thought to be critical for fast Li+ transport. 16, 18 The doping may alter the vacancy distribution in the octahedral site to impact the conductivity. With this hypothesis, a question is then raised: is there a rule to control the vacancy distribution by aliovalent doping to trigger the fast Li+ conduction?

In this work, we chose two Li-site-doping garnets by two distinct dopants (Al, Zn): Li6. 28Al0. 24La3Zr2O12 (abbr. LLZO-Al24) and Li5. 8Zn0. 6La3Zr2O12 (abbr. LLZO-Zn60), along with LLZO as reference. The two dopants belong to different families and groups, have different contents and valences, and bring different amounts of vacancies. We then discerned a common rule that the dopants obey in redistributing the vacancies in the framework of the garnets. The Li vacancy density in the active