The paper studies a class of graphs called 1-planar graphs, which are graphs that can be drawn on a flat surface so that each edge is crossed by at most one other edge. Within this class, the authors look for the graph that maximizes the spectral radius, a number derived from the graph's adjacency matrix that captures how densely and evenly connected the graph is. The additional constraint is that the graph must not contain t or more separate copies of a five-cycle (a pentagon) as subgraphs. Finding such extremal graphs is a classical type of problem in combinatorics, sitting at the intersection of graph theory and linear algebra.
The main result identifies, for every fixed t of at least 3 and for all large enough graphs, the unique graph in this class that achieves the highest spectral radius. The authors show that any graph achieving this maximum must have a very specific structure: after removing two highly connected "dominating" vertices, what remains is built almost entirely from repeated copies of a particular small graph on seven vertices, plus at most one additional bounded piece. To establish this, they prove a key inequality relating the number of vertices, edges, and pentagons in any candidate graph. This inequality tightly controls how much the structure can deviate from the idealized repeating pattern.
The proof strategy then uses a resolvent technique from spectral graph theory to eliminate the repeated building blocks algebraically, and applies moment comparisons to pin down the remaining piece uniquely depending on the size of the graph modulo 7. The boundary case of exactly three pentagons requires a separate careful argument. The work resolves an open problem posed by Li, Wang, and Zhao, and illustrates how combining structural graph theory with spectral methods can yield precise, unique answers to extremal questions.