Abstract:In this paper we introduce geometric crystals and unipotent crystals which are algebro-geometric analogues of Kashiwara's crystal bases. Given a reductive group G, let I be the set of vertices of the Dynkin diagram of G and T be the maximal torus of G. The structure of a geometric G-crystal on an algebraic variety X consists of a rational morphism \gamma:X-->T and a compatible family e_i:G_m\times X-->X, i\in I of rational actions of the multiplicative group G_m satisfying certain braid-like relations. Such a structure induces a rational action of W on X. Quite surprisingly, many interesting rational actions of the group W come from geometric crystals. Also all the known examples of the action of W which appear in the construction of gamma-functions for the representations of ^LG in the recent work by A. Braverman and D. Kazhdan come from geometric crystals. There are many examples of positive geometric crystals on (G_m)^l, i.e., those geometric crystals for which the actions e_i and the morphism \gamma are given by positive rational expressions. To each positive geometric crystal X we associate a Kashiwara's crystal corresponding to the Langlands dual group ^LG. An emergence of ^LG in the "crystal world" was observed earlier by G. Lusztig. Another application of geometric crystals is a construction of trivialization which is an W-equivariant isomorhism X-->\gamma^{-1}(e) \times T for any geometric SL_n-crystal. Unipotent crystals are geometric analogues of normal Kashiwara crystals. They form a strict monoidal category. To any unipotent crystal built on a variety X we associate a certain gometric crystal.
What problem does this paper attempt to address?
The problems that this paper attempts to solve mainly focus on the research of geometric crystals and unipotent crystals. Specifically, the author attempts to study the structures related to Kashiwara's crystal bases by means of algebraic geometry and explore the applications of these structures in Lie groups and their representation theories.
### Main problems of the paper include:
1. **Definition and properties of geometric crystals**:
- The author introduced the concept of geometric crystals and studied their basic properties. Geometric crystals are the algebraic - geometric analogues of crystal bases, which consist of an algebraic variety \( X \) and some specific rational maps.
- Defined the Weyl group action on geometric crystals and proved that many interesting rational actions of the Weyl group originate from geometric crystals.
2. **Introduction of unipotent crystals**:
- Unipotent crystals are another algebraic - geometric analogue of crystal bases, especially suitable for the crystal bases of \( U_q(L\mathfrak{g}) \)-modules. Unipotent crystals consist of an irreducible algebraic variety \( X \) and a rational action of a one - parameter subgroup \( U \).
- Studied the product structure of unipotent crystals and proved that the category of unipotent crystals is a strictly monoidal category under this product.
3. **Geometrization and tropicalization**:
- Introduced the concept of "tropicalization", that is, a set of formal rules from birational isomorphisms to piecewise linear maps. Through tropicalization, geometric crystals can be transformed into combinatorial crystals.
- Studied the trivialization of geometric crystals and gave explicit formulas in some special cases.
4. **Research on standard unipotent crystals**:
- Studied standard unipotent crystals, especially their orbit decompositions in the quotient spaces of Borel subgroups.
- Constructed \( W \)-invariant functions on some standard unipotent crystals, which are very important for the study of the \( \gamma \)-functions of representations.
5. **Connections with other works**:
- Connected unipotent crystals with the work of Braverman and Gaitsgori, who constructed crystal bases by perverse sheaves on the affine Grassmannian.
- Explored the relationship between geometric crystals and the Langlands dual group \( L G \), and found that \( L G \) plays an important role in the "crystal world".
### Summary
The core problem of this paper is to study geometric crystals and unipotent crystals by means of algebraic geometry, aiming to understand the significance of these structures in Lie groups and their representation theories and explore their relationships with combinatorial crystals, \( \gamma \)-functions and other mathematical objects.