Zero-shot Object Navigation using Vision-Language Models with Tree of Thoughts Reasoning