๐”ฐ๐”ฉ2โก๐•‚

Affine Lie algebras of ๐”ฐ๐”ฉ2โก๐•‚

Let ๐”ž = ๐”ฐ๐”ฉ2โก๐•‚ with its Chevalley basis and let ๐œŽ๐‘– be the suggestively named1 involutive isometries of ๐”ž defined by #m/def/lie

๐œŽ1:๐›ผ1โ†ฆโˆ’๐›ผ1๐œŽ2:๐›ผ1โ†ฆโˆ’๐›ผ1๐œŽ3:๐›ผ1โ†ฆ๐›ผ1๐‘ฅยฑ๐›ผ1โ†ฆ๐‘ฅโˆ“๐›ผ1๐‘ฅยฑ๐›ผ1โ†ฆโˆ’๐‘ฅโˆ“๐›ผ1๐‘ฅยฑ๐›ผ1โ†ฆโˆ’๐‘ฅยฑ๐›ผ

and let ๐œŽ0 =1. Furthermore we let

๐‘ฅ+๐›ผ1=๐‘ฅ๐›ผ1+๐‘ฅโˆ’๐›ผ1=๐‘ฅ๐›ผ1+๐‘ฅ๐œŽ1๐›ผ1=๐‘ฅ๐›ผ1โˆ’๐‘ฅ๐œŽ2๐›ผ1๐‘ฅโˆ’๐›ผ1=๐‘ฅ๐›ผ1โˆ’๐‘ฅโˆ’๐›ผ1=๐‘ฅ๐›ผ1โˆ’๐‘ฅ๐œŽ1๐›ผ1=๐‘ฅ๐›ผ1+๐‘ฅ๐œŽ2๐›ผ1

We consider the untwisted or twisted affine Lie algebra ห†๐”ž๐‘– =ห†๐”ž[๐œŽ๐‘–]2 which have bases

ห†๐”ž0=โŸจ๐‘,๐›ผ1โŠ—๐‘ก๐‘š,๐‘ฅยฑ๐›ผ1โŠ—๐‘ก๐‘š:๐‘šโˆˆโ„คโŸฉห†๐”ž1=โŸจ๐‘,๐›ผ1โŠ—๐‘ก๐‘š+1/2,๐‘ฅ+๐›ผ1โŠ—๐‘ก๐‘š,๐‘ฅโˆ’๐›ผ1โŠ—๐‘ก๐‘š+1/2:๐‘šโˆˆโ„คโŸฉห†๐”ž2=โŸจ๐‘,๐›ผ1โŠ—๐‘ก๐‘š+1/2,๐‘ฅ+๐›ผ1โŠ—๐‘ก๐‘š+1/2,๐‘ฅโˆ’๐›ผ1โŠ—๐‘ก๐‘š:๐‘šโˆˆโ„คโŸฉห†๐”ž3=โŸจ๐‘,๐›ผ1โŠ—๐‘ก๐‘š,๐‘ฅยฑ๐›ผ1โŠ—๐‘ก๐‘š+1/2:๐‘šโˆˆโ„คโŸฉ

The 1-dimensional subalgebra ๐”ฅ =๐•‚๐›ผ โ‰ค๐”ž generates the natural Heisenberg algebras

ห†๐”ฅโ„ค=๐•‚๐‘โŠ•โจ๐‘›โˆˆโ„คโˆ–{0}๐›ผ1โŠ—๐‘ก๐‘›โ‰คหœ๐”ฅโ‰คห†๐”ž0,ห†๐”ž3ห†๐”ฅโ„ค+12=๐•‚๐‘โŠ•โจ๐‘›โˆˆโ„ค+12๐›ผ1โŠ—๐‘ก๐‘›โ‰คหœ๐”ฅ[โˆ’1]โ‰คห†๐”ž1,ห†๐”ž2

as Lie subalgebras and we have3

[๐‘,ห†๐”ž๐‘–]=0[๐›ผ1โŠ—๐‘ก๐‘š,๐›ผ1โŠ—๐‘ก๐‘›]=2๐‘š๐›ฟ๐‘š+๐‘›๐‘
Isomorphism of extended Lie algebras

หœ๐”ž and หœ๐”ž[๐œŽ3] are isomorphic (but not as graded Lie algebras) under

๐‘โ†ฆ๐‘๐‘‘โ†ฆ๐‘‘โˆ’14๐›ผ1๐›ผ1โŠ—๐‘ก๐‘›โ†ฆ๐›ผ1โŠ—๐‘ก๐‘›+12๐›ฟ๐‘›๐‘๐‘ฅยฑ๐›ผ1โ†ฆ๐‘ฅยฑ๐›ผ1โŠ—๐‘ก๐‘›ยฑ1/2

There also exist grade-preserving isomorphisms between หœ๐”ž[๐œŽ๐‘–] for ๐‘– =1,2,3.

Via formal series

Taking a formal series approach on ห†๐”ž๐‘–, the exact characterization varies with ๐‘–.

For ห†๐”ž0

In ห†๐”ž0[[๐‘ง,๐‘งโˆ’1]] we define the formal sums

๐‘ฅยฑ๐›ผ1(๐‘ง)=โˆ‘๐‘›โˆˆโ„ค(๐‘ฅยฑ๐›ผ1โŠ—๐‘ก๐‘›)๐‘งโˆ’๐‘›๐›ผ1(๐‘ง)=โˆ‘๐‘›โˆˆโ„ค(๐›ผ1โŠ—๐‘ก๐‘›)๐‘งโˆ’๐‘›

the commutation relations are more conveniently expressed as

[๐›ผ1โŠ—๐‘ก๐‘š,๐‘ฅยฑ๐›ผ1(๐‘ง)]=ยฑ2๐‘ง๐‘š๐‘ฅยฑ๐›ผ1(๐‘ง)=โŸจ๐›ผ1,ยฑ๐›ผ1โŸฉ๐‘ง๐‘š๐‘ฅยฑ๐›ผ1(๐‘ง)[๐‘ฅยฑ๐›ผ1(๐‘ง1),๐‘ฅยฑ๐›ผ(๐‘ง2)]=0[๐‘ฅ๐›ผ1(๐‘ง1),๐‘ฅโˆ’๐›ผ1(๐‘ง2)]=(๐›ผ1(๐‘ง2)โˆ’๐‘๐ท1)๐›ฟ(๐‘ง1/๐‘ง2)[๐‘‘,๐‘ฅยฑ๐›ผ1(๐‘ง)]=โˆ’๐ท๐‘ฅยฑ๐›ผ1(๐‘ง)[๐‘‘,๐›ผ1(๐‘ง)]=โˆ’๐ท๐›ผ1(๐‘ง)

where ๐‘š โˆˆโ„ค.

For ห†๐”ž3

In ห†๐”ž3[[๐‘ง1/2,๐‘งโˆ’1/2]] we define the formal sums

๐‘ฅยฑ๐›ผ1(๐‘ง)=โˆ‘๐‘›โˆˆโ„ค+12(๐‘ฅยฑ๐›ผ1โŠ—๐‘ก๐‘›)๐‘งโˆ’๐‘›๐›ผ1(๐‘ง)=โˆ‘๐‘›โˆˆโ„ค(๐›ผ1โŠ—๐‘ก๐‘›)๐‘งโˆ’๐‘›

the commutation relations are more conveniently expressed as

[๐›ผ1โŠ—๐‘ก๐‘š,๐‘ฅยฑ๐›ผ1(๐‘ง)]=ยฑ2๐‘ง๐‘š๐‘ฅยฑ๐›ผ1(๐‘ง)=โŸจ๐›ผ1,ยฑ๐›ผ1โŸฉ๐‘ง๐‘š๐‘ฅยฑ๐›ผ1(๐‘ง)[๐‘ฅยฑ๐›ผ1(๐‘ง1),๐‘ฅยฑ๐›ผ(๐‘ง2)]=0[๐‘ฅ๐›ผ1(๐‘ง1),๐‘ฅโˆ’๐›ผ1(๐‘ง2)]=(๐›ผ1(๐‘ง2)โˆ’๐‘๐ท1)[(๐‘ง1/๐‘ง2)1/2๐›ฟ(๐‘ง1/๐‘ง2)][๐‘‘,๐‘ฅยฑ๐›ผ1(๐‘ง)]=โˆ’๐ท๐‘ฅยฑ๐›ผ1(๐‘ง)[๐‘‘,๐›ผ1(๐‘ง)]=โˆ’๐ท๐›ผ1(๐‘ง)
For ห†๐”ž1

In ห†๐”ž1[[๐‘ง1/2,๐‘งโˆ’1/2]] we define the formal sums

๐‘ฅยฑ๐›ผ1(๐‘ง)=12โˆ‘๐‘›โˆˆโ„ค(๐‘ฅ+๐›ผ1โŠ—๐‘ก๐‘›)๐‘งโˆ’๐‘›ยฑ12โˆ‘๐‘›โˆˆโ„ค+12(๐‘ฅโˆ’๐›ผ1โŠ—๐‘ก๐‘›)๐‘งโˆ’๐‘›๐›ผ1(๐‘ง)=โˆ‘๐‘›โˆˆโ„ค+12(๐›ผ1โŠ—๐‘ก๐‘›)๐‘งโˆ’๐‘›

the commutation relations are more conveniently expressed as

[๐›ผ1โŠ—๐‘ก๐‘š,๐‘ฅยฑ๐›ผ1(๐‘ง)]=ยฑ2๐‘ง๐‘š๐‘ฅยฑ๐›ผ1(๐‘ง)=โŸจ๐›ผ1,ยฑ๐›ผ1โŸฉ๐‘ง๐‘š๐‘ฅยฑ๐›ผ1(๐‘ง)[๐‘ฅ๐›ผ1(๐‘ง),๐‘ฅโˆ’๐›ผ1(๐‘ง2)]=12(๐›ผ1(๐‘ง2)โˆ’๐‘๐ท1)๐›ฟ(๐‘ง1/21/๐‘ง1/22)

and we also have

๐‘ฅโˆ’๐›ผ1(๐‘ง)=lim๐‘ง1/2โ†’โˆ’๐‘ง1/2๐‘ฅ๐›ผ1(๐‘ง)

Representations


#state/develop | #lang/en | #SemBr

Footnotes

  1. For ๐•‚ =โ„‚, we can conjugate by Pauli matrices ๐œŽ๐‘– for the same result. โ†ฉ

  2. FLM use ๐œ—1 =๐œŽ3 and ๐œ—2 =๐œŽ1 โ†ฉ

  3. 1988. Vertex operator algebras and the Monster, ยง3.1, pp. 62โ€“67 โ†ฉ