Self-adjoint element

In mathematics, an element of a *-algebra is called self-adjoint if it is the same as its adjoint (i.e. a=a∗).

Definition

Let 𝒜 be a *-algebra. An element a∈𝒜 is called self-adjoint if a=a∗.

The set of self-adjoint elements is referred to as 𝒜sa.

A subset ℬ⊆𝒜 that is closed under the involution *, i.e. ℬ=ℬ∗, is called self-adjoint.

A special case of particular importance is the case where 𝒜 is a complete normed *-algebra, that satisfies the C*-identity (‖a∗a‖=‖a‖2 ∀a∈𝒜), which is called a C*-algebra.

Especially in the older literature on *-algebras and C*-algebras, such elements are often called hermitian. Because of that the notations 𝒜h, 𝒜H or H(𝒜) for the set of self-adjoint elements are also sometimes used, even in the more recent literature.

Examples

Criteria

Let 𝒜 be a *-algebra. Then:

Properties

In *-algebras

Let 𝒜 be a *-algebra. Then:

  • Each element a∈𝒜 can be uniquely decomposed into real and imaginary parts, i.e. there are uniquely determined elements a1,a2∈𝒜sa, so that a=a1+ia2 holds. Where a1=12(a+a∗) and a2=12i(a−a∗).
  • The set of self-adjoint elements 𝒜sa is a real linear subspace of 𝒜. From the previous property, it follows that 𝒜 is the direct sum of two real linear subspaces, i.e. 𝒜=𝒜sa⊕i𝒜sa.
  • If a∈𝒜sa is self-adjoint, then a is normal.
  • The *-algebra 𝒜 is called a hermitian *-algebra if every self-adjoint element a∈𝒜sa has a real spectrum σ(a)⊆ℝ.

In C*-algebras

Let 𝒜 be a C*-algebra and a∈𝒜sa. Then:

  • For the spectrum ‖a‖∈σ(a) or −‖a‖∈σ(a) holds, since σ(a) is real and r(a)=‖a‖ holds for the spectral radius, because a is normal.
  • According to the continuous functional calculus, there exist uniquely determined positive elements a+,a−∈𝒜+, such that a=a+−a− with a+a−=a−a+=0. For the norm, ‖a‖=max⁡(‖a+‖,‖a−‖) holds. The elements a+ and a− are also referred to as the positive and negative parts. In addition, |a|=a++a− holds for the absolute value defined for every element |a|=(a∗a)12.
  • For every a∈𝒜+ and odd n∈ℕ, there exists a uniquely determined b∈𝒜+ that satisfies bn=a, i.e. a unique n-th root, as can be shown with the continuous functional calculus.

See also

Notes

  1. 1 2 3 4 5 6 Dixmier 1977, p. 4.
  2. ↑ Dixmier 1977, p. 3.
  3. ↑ Palmer 2001, p. 800.
  4. ↑ Dixmier 1977, pp. 3–4.
  5. 1 2 Kadison & Ringrose 1983, p. 271.
  6. ↑ Palmer 2001, pp. 798–800.
  7. ↑ Palmer 2001, p. 798.
  8. ↑ Palmer 2001, p. 1008.
  9. ↑ Kadison & Ringrose 1983, p. 238.
  10. ↑ Kadison & Ringrose 1983, p. 246.
  11. ↑ Dixmier 1977, p. 15.
  12. ↑ Blackadar 2006, p. 63.

References