find and pivot substitutes, Sql Server

Language: Layout:

DECLARE @table1 TABLE
(
    [Sub1] CHAR(1),
    [Sub2] CHAR(1)
);

INSERT INTO @table1
VALUES
('A', 'B'),
('A', 'D'),
('A', 'F'),
('A', 'E'),
('A', 'A'),
('A', 'C'),
('A', 'G'),
('C', 'A'),
('B', 'A'),
('X', 'Z');

declare @DoubleSub table (
    [Sub1] CHAR(1) NOT NULL,
    [Sub2] CHAR(1) NOT NULL,
    PRIMARY KEY ([Sub1], [Sub2])
);

INSERT INTO @DoubleSub ([Sub1], [Sub2])
SELECT [Sub1], [Sub2] FROM @table1
UNION
SELECT [Sub2], [Sub1] FROM @table1;

;WITH TRIPLESUB AS
(
    SELECT [Sub1] as [Main], 1 as Lvl, [Sub1], [Sub2], CAST('|'+ [Sub1] +'|'+ [Sub2] +'|' AS VARCHAR(200)) as Visited
    FROM @DoubleSub
    
    UNION ALL
    
    SELECT c.[Main], c.Lvl + 1, t.[Sub1], t.[Sub2],
    CAST(c.Visited + t.[Sub2] + '|' AS VARCHAR(200))
    FROM TRIPLESUB c
    JOIN @DoubleSub t
      ON t.[Sub1] = c.[Sub2] 
     AND c.Lvl < 6
     AND c.[Main] != t.[Sub2]
     AND c.Visited NOT LIKE CONCAT('%|',t.[Sub2],'|%')
)
SELECT 
[Main], 
[1] as [Substitute1],
[2] as [Substitute2],
[3] as [Substitute3],
[4] as [Substitute4],
[5] as [Substitute5],
[6] as [Substitute6]
FROM 
(
  SELECT [Main], [Sub2],
   ROW_NUMBER() OVER (PARTITION BY [Main] ORDER BY IIF([Main]<[Sub2],0,1), [Sub2]) AS RN
  FROM TRIPLESUB
  WHERE [Main] != [Sub2]
  GROUP BY [Main], [Sub2]
) AS src
PIVOT
(
  MAX([Sub2])
  FOR RN IN ([1],[2],[3],[4],[5],[6])
) AS pvt;

View schema
Execution time: 0,06 sec, rows selected: 9, rows affected: 25, absolute service time: 0,36 sec, absolute service time: 0,4 sec