Emotions are a key aspect of being human. When people write essays or reports, they naturally express their own feelings. Recently, the way we identify emotions in text has been completely transformed by using transformer-based models in natural language processing (NLP). Transformer-based models have revolutionized text emotion detection through Contextual Understanding, Long-range Dependencies, Transfer Learning, Multi-modal Inputs, and Fine-tuning. This review paper thoroughly investigates the present status of investigations using transformers to recognize and understand emotions conveyed in written text. We examine several studies that have utilized transformer architectures, starting from the influential Transformer model to more recent variations like BERT, GPT, XLM, and their derived models. This review paper begins by introducing various emotion models. Following that, it discusses the process of detecting emotions in text. The paper explores different types of embeddings and provides a detailed discussion on transformers and their variations. The literature review section critically assesses existing methodologies, identifying trends, advancements, and gaps in current knowledge. Results from various studies are synthesized and compared. This comprehensive review aims to serve as a valuable asset for scholars, practitioners, and enthusiasts interested in the intersection of transformers and text emotion detection. Moreover, the paper examines the challenges faced in the domain of text emotion detection.