The Impact of Negated Text on Hallucination with Large Language Models