replace() leaves unicode variation selector-16

[ogslumber]

Describe the bug replace() function leaves unicode variation selector-16 (\xef\xb8\x8f) when replacing Repeat Button emoji (🔁️).

To Reproduce

import demoji

sample_var = '🔁️ sample text'
print(sample_var.encode('utf-8'))
>>> b'\xf0\x9f\x94\x81\xef\xb8\x8f sample text'

sample_var = demoji.replace(sample_var)
print(sample_var.encode('utf-8'))
>>> b'\xef\xb8\x8f sample text'

Expected behavior String without \xef\xb8\x8f sequence:

>>> b' sample text'