here are simple steps

1. Export XML: Use the to_xml() function as usual.
2. Post-Process: Open the file, replace / with xsi:nil="true", and save it.
3. Output: The updated XML will correctly show <data_grafic xsi:nil="true" />

Unfortunately, the to_xml() method in pandas does not currently support directly adding attributes like xsi:nil="true" for null values. However, you can achieve your desired result by post-processing the XML output. After exporting the DataFrame to XML, modify the output string to replace empty elements with the desired format:

import pandas as pd

# Example DataFrame
data = {'col1': [1, None], 'col2': [None, 4]}
df = pd.DataFrame(data)

# Export to XML
xml_output = df.to_xml(
    'df.xml',
    root_name='dataroot',
    row_name='Anexa_1I',
    namespaces={
        "xsd": "http://www.w3.org/2001/XMLSchema",
        "xsi": "http://www.w3.org/2001/XMLSchema-instance",
    },
    prefix="",
    xml_declaration=False,
    index=False,
    pretty_print=True,
)

# Post-process XML file to add `xsi:nil="true"` for null values
with open('df.xml', 'r') as file:
    xml_content = file.read()

# Replace self-closing empty elements with xsi:nil="true"
updated_xml = xml_content.replace('/>', ' xsi:nil="true" />')

# Save the updated XML
with open('df.xml', 'w') as file:
    file.write(updated_xml)

print("XML updated with xsi:nil='true' for null values.")

Even we have different blogs that can be referred for further help- https://techlusion.io/insight/