pandas - convert string into many columns

data = [
    'TTMNPMGN=31.28954;NLOW=307.97567;TTMPRCFPS=27.16872',
    'NLOW=307.97567;TTMPRCFPS=27.16872;TTMGROSMGN=49.10444',
]

import pandas as pd

print('\n--- original data ---\n')

df1 = pd.DataFrame(data)
print(df1.head())

print('\n--- splitted to columns ---\n')

df2 = pd.DataFrame()

for index, row in df1.iterrows():
    d = dict(item.split('=') for item in row[0].split(';'))
    df2 = df2.append(d, ignore_index=True)

print(df2.head())

print('\n--- filled missing values ---\n')

#df2.fillna('0', inplace=True)
df2['TTMNPMGN'].fillna('999', inplace=True)
df2['TTMGROSMGN'].fillna('-999', inplace=True)
print(df2.head())

# save using "columns" to keep order of columns in file
df2.to_csv('output.csv', index=False, columns=['TTMNPMGN', 'NLOW', 'TTMPRCFPS', 'TTMGROSMGN'])

'''
--- original data ---

                                                   0
0  TTMNPMGN=31.28954;NLOW=307.97567;TTMPRCFPS=27....
1  NLOW=307.97567;TTMPRCFPS=27.16872;TTMGROSMGN=4...

--- splited to columns ---

        NLOW  TTMNPMGN TTMPRCFPS TTMGROSMGN
0  307.97567  31.28954  27.16872        NaN
1  307.97567       NaN  27.16872   49.10444

--- filled missing values ---

        NLOW  TTMNPMGN TTMPRCFPS TTMGROSMGN
0  307.97567  31.28954  27.16872       -999
1  307.97567       999  27.16872   49.10444
'''