How can I flatten this json and create a table with the id presented in the table ?

Json file structure as below is present as a string in a large table

 [
      {
        "a": {
          "b": [
            {
              "c": "1",
              "d": "2",
              "e": "3",
              "f": "4"
            }
          ]
        },
        "g": {
          "h": "5",
          "i": "6"
        }
      }
    ]

I have tried below yet it doens't seem to split it correctly.

 pd.DataFrame(pd.json_normalize(df,record_path=["a","b"],["g"])) 

I am expecting a data frame that i can work with