The codes on the page are taken directly from the article text. If you view the source HTML on your page, you'll see the following:
<p style="text-align: justify;"><span style="background-image: initial; background-position: initial; background-repeat: initial; background-attachment: initial; font-family: arial, helvetica, sans-serif; font-size: 12pt;">O Jáqtáqvá/Friolack bateu o Cometa/Rodeio Futsal, neste sábado, pela 2ª rodada da Chave 8 do Estadual de Futsal Série Bronze 2017. Jogando em casa, o alvi-roxo venceu por 6 a 3, com gols de Tininho, Jarré(2x) e o artilheiro Andrei Marcolan(3x). Com o triunfo, a equipe chegou a 6 pontos se isolando na ponta da tabela seguido por a AMF de Marau que também venceu seu jogo em casa e garantiu seus primeiros 3 pontos na tabela.</span></p>
We use that text in the Open Graph description.
You may want to check that your database is properly storing the text in UTF-8 encoding because, from the above, it doesn't look like it. Browsers, and even Facebook, can decipher those escape code sequences to the proper character. I think the issue here is that Facebook simply hadn't scanned and processed your page yet, so the output was the raw text. If you use Facebook URL debug tool, you'll see the description looks correct now:
developers.facebook.com/tools/debug/shar...e-expuls%25C3%25B5es
You'll want to figure out why the right characters aren't in the actual article first. Otherwise, you may need to use the Debug Tool to have Facebook scan your page first before you share it out.
I hope that helps explain,
Alex
Encoding og:description
