Large language models (LLMs) are gaining popularity in different aspects of real-world design problems, including design-ideation. While there is significant potential for using LLMs for concept generation, there is a lack of knowledge about how LLMs can be effectively used to maximize the diversity of generated design concepts. In this study, we hypothesize that LLMs can be used to generate a diverse set of design concepts by prompting with different personas that represent professional identities. We test this hypothesis by generating multiple knowledge bases using the topics from different professional personas and providing prompts to ChatGPT that are aligned with these personas. Our results indicate that ChatGPT can generate design concepts that are more semantically diverse by providing prompts with different professional personas, compared to providing prompts without specifying any persona. We conclude that LLMs, such as ChatGPT, can facilitate design ideation and provide increased diversity of concepts through professional personas.