AFIK，这不能用Pandas来做。必须使用openpyxl。
下面是基于您的主file的itertools.groupby的建议。

from openpyxl import load_workbook
from itertools import groupby
wb = load_workbook("myfile.xlsx")
ws = wb.active # or wb["Sheet2"]
data = ws.rows
header = [(h._style, h.value) for h in next(data)]
pairs = [[(r._style, r.value) for r in rows] for rows in data]
groups = {
    std: list(g) for std, g in groupby(
        sorted(pairs, key=lambda x: x[1][1]), key=lambda x: x[1][1])
}
for std, vals in groups.items():
    std_ws = wb.create_sheet(std)
        
    for idxr, rows in enumerate(*[[header] + vals], 1):
        for idxc, (s, v) in enumerate(rows, 1):
            wb[std].cell(idxr, idxc).value = v
            wb[std].cell(idxr, idxc)._style = s
    # https://stackoverflow.com/a/52736133/16120011
    from copy import copy
    for idx, rd in ws.column_dimensions.items():
        std_ws.column_dimensions[idx] = copy(rd)
del wb["Sheet2"] # could be unnecessary
wb.save("newfile.xlsx")

输出量：