Privacy-Aware Semantic Cache for Large Language Models