We categorized clumps, embedded clusters, and open clusters and conducted a comparative analysis of their physical properties.
Overall, the radii of open clusters are significantly larger than those of embedded clusters and clumps. The radii of embedded clusters are larger than those of clumps, which may be due to the expansion of embedded clusters.
The open clusters have significantly higher masses than embedded clusters, by about one order of magnitude.
Given the current mass distribution of clumps in the Milky Way, the evolutionary sequence from a single clump evolving into an embedded cluster and subsequently into an open cluster cannot account for the observed open clusters with old ages and high masses, which is also supported by N-body simulations of individual embedded clusters.
To explain the mass and radius distributions of the observed open clusters, initial embedded clusters with masses higher than 3000 M$_ odot $ are necessary.
However, the upper limit of the embedded cluster sample is less than 1000 M$_ odot $, and only a few ATLASGAL clumps have a mass higher than 3000 M$_ odot $.
Thus, the currently observed clumps cannot be the "direct" precursors of the currently observed open clusters.
If the Milky Way has a burst-like and time-dependent star formation history, the currently observed open clusters with old ages and high masses may come from massive clumps in the past. There is also a very real possibility that these open clusters originate from post-gas expulsion coalescence of multiple embedded clusters. We compared the separation of open clusters and the typical size of molecular clouds, and find that most molecular clouds may only form one open cluster, which supports the scenario of post-gas expulsion coalescence. Further study is necessary to distinguish between the different scenarios.