❔ Splittings UT8 text

I have an UTF8 text file text that is delimited by ╚ for lines and then further by & for values. I read all text with UTF8 enconding. I split it first by main delimiter, then by secondary. I do basic processing, no conversions. I serialize it with System.Text.Json via Encoder = JavaScriptEncoder.Create(UnicodeRanges.BasicLatin, UnicodeRanges.Cyrillic I write it back to text file with UTF8 encoding. And yet I get this picrelated. Since I operate with string objects there is little to no chance I somehow split in the middle of grapheme or unicode symbol. So what gives?