It is also common to use two different samples for the two or more stages. Your answer lies in the purpose. If researching a small, esoteric sample or only one researcher, it is probably too difficult or impractical to secure another sample. Another issue, probably most common, is the same sample is of interest throughout the study. When two or more samples are used, the sample size is larger, available, and often used for developing scales or greater generalizability/transferability. For example, focus groups to develop a scale and then piloting the scale with a large sample for statistical analysis. Another example would be qualitative analysis following a questionnaire of multiple groups.

My book, Integrating Qualitative and Quantitative Methods, has a chapter on this design (QUAN --> qual), and it includes support for the idea that you do not have to rely on participants from the original quantitative study in the follow-up qualitative study.

I have found your book Professor David L Morgan , thank you very much for the recommendation. I will go through the discussion from your book. Thanks also to Dr. David Coker for your point of view, its helpfull. I found a new thing about what you have said that fits with my evaluation study in social science which is called Multi-level Mix Method Research Design. This design uses different levels of population and also uses an integration approach in terms of structure, components and mechanism. I think its more suitable design for me compared to sequential explanatory. https://journals.sagepub.com/doi/abs/10.1177/1558689819844417