Usually when I get enough of something new (a.k.a. “theoretical” discoveries), I start to think about how to put them into paper(s). Since different journals have different styles/directions, I also need to decide which journal to target. I think research results can be divided into the following categories:

1. Introducing new concepts worth enough for whole science/physics community to know.
2. Solving a puzzle that has been around and popular quite for a while.
3. (a) Important breakthrough/progress in a specialized field which needs attention from the whole community.
   (b) Same as (a) but relevant only within a specific community (but still a very important piece of work).
4. Providing thorough analysis and extensive data although the idea or main result is known already.

Though it is my personal impression, but here are some examples. For categories 1 and 2, most people seem to submit to Nature journals. Category 3(a) is for PRL. Category 3(b) is a bit in a wierd position. I would go for PRL but it’s 50-50 chance rejected. It actually depends a lot on the luck, which referees you encounter, rather than actual contents of the paper. Category 4 is for Physical Review specialized joruanls (e.g. PRB, PR Materials, PR Applied, PR Research)

My research projects are quite diverse, but mostly the results, especially orbitronics projects, are within categories 3 and 4. Over the last 2 years, I tried various top journals for publishing experiment + theory papers together with experimentalists, which are mostly about new findings in orbitronics research. But at least so far, our papers have been all rejected. Actually, most rejections were from editorial screening level even before indepth reviews. I thought our results are really important, but I think most researchers would think in the same way. Anyway, it’s a pity.

When I write papers of my own as a leading author, I think of the following criterion: Is this paper going to be still read after decades? In spintronics, I believe Berger/Slonczewski’s theoretical proposal of the concept of spin-transfer torque and a thorough analysis by Stiles and Zangwill are famous examples, which correspond to category 3 and 4, respectively. Hypothetically speaking, if they publish their papers now, I don’t really see how their papers appear in prestigious journals. Nonetheless, these paper are truely excellent, which have provided, and still playing, a pivotal role in the whole spintronics research community.

After all, I wouldn’t really mind that I cannot publish in top journals as long as there are enough people who appreciate contents of my paper. Recently I was told from a few colleagues that they are learning a lot from my PR Research paper, which made me really happy. Well, actually that was my purpose: To write a paper from which somebody can actually “learn” rather than catching pieces of key messages. But for postdocs, making a CV more attractive is necessary, and one cannot deny that publications from prestigious journals catch hiring committee’s eyes. It’s similar situation for application and evaluation of research grants. Probably that’s why there’s high competition in publishing, which seems to be more than necessary. I dont’ think it’s good for science culture at all. In general, the pluralism is essential to build more stable and reliable knowledge system. The puralism is definitely not possible if everybody does similar research works as papers published in top journals.

PS. Maybe that’s why I like APS journals. For most papers in specialized journals such as PRB, they are not cited much in the beginning (which is why the IF is not high) but some of them are cited significantly as time goes on.