interesting (Why should other people think that your work is important?)
at the proper level (You know all about it, but other people do not.)
Content should contain the relevant material and should not contain seemingly irrelevant material.
The paper should be well structured.
Title should be concise, but should describe what you have done accurately.
Abstract should be a short summary of your work, but should be accurate and interesting (many people will not read the rest of the paper if they do not find the abstract interesting).
Introduction should describe why you have worked on this topic and it's relevance to previous work. You should give a lot of references to work of others so that an interested reader may find out more about the problem by reading these references. Do not cite work you yourself have not looked at. If that paper was important, you should have studied it!
The next part should give a more detailed description of the theory of what you did. You should put the equations, calculations, etc. which relate to what is expected from your work. Clearly indicate whose results or calculations you are using. (You should not give the details of a calculation but just the main steps.)
The next part should contain the main part of the work you did. Again, clearly indicate what you did.
The last section should summarize your results and what it means in relation to previous work. Discuss the difficulties you had, and what could be done in the future to understand the problem better.
Sometimes you need to add some extra material of technical nature, such as the details of a calculation, or computer code. These should go into an appendix.
References should be consistent in style. All references to journals should look the same, similarly references to books, etc. Check APS papers if you want to see examples.
Be precise. Tell the exact conditions your results were obtained. If you made an observation that sometimes occurred and sometimes did not, it means you either did not control a variable that you should have, or that the process has some randomness in it, which should be quantified.
Some notes on grammar:
Use third person - The experiment was carried out in place of I carried out the experiment.
Do not use contractions, use is not in place of isn't
Do not use copyrighted material in your manuscript without permission. Some web pages will allow you to use the material, but will ask you to acknowledge them in a certain specific way.