[reportlab-users] Problem with some (not all) unicode symbols

[Бачо Киро]

My python code creates tables in PDF file. In header rows there are unicode symbols. Some of them are OK, but another are like square.

I use fonts Verdana and VerdanaB:

pdfmetrics.registerFont(TTFont('Verdana', 'Verdana.ttf'))
pdfmetrics.registerFont(TTFont('VerdanaB', 'VerdanaB.ttf'))

Texts like mg/Nm³ and °C are visualized correctly. But texts like SO₂ and NOₓ does not. They look so: SO◻ and NO◻.
How I can fix it?