A good topic has a strong, clear "take home message". The bulk of the paper/talk must explain the take home message. Science explanation must work for non-science (paper) or science (talk) audience.
Consider "bar code scanners". Science is how a scanner reads the code, independent of its orientation. But how the computer, having read the bar code, looks up the UPC in a database so the paper concerns databases and not bar codes.
A poor topic is quantum computing, if paper goes like:
"quantum computers will avoid the one-atom limit (false),
uses qubits for data storage and computation,
can do true multitasking (misleading) and
has applications in cryptography."
How quantum-specific algorithms work is a difficult and an uninteresting.
Topics need to have some depth. Consider the topic "how fireworks produce different colors". If the content is limited to a brief discussion of emission lines (for the non-science audience) and a list of which chemicals produce which colors, the content will insufficient to interest your science colleagues.
Astronomy topics are difficult as science tends to be vague.
History of science works only if history illustrates the science.
For papers and talks, the topic must contain both an idea that can be explained to a non-technical audience and enough hard, new science to be interesting to physics majors.
For non-technical audience, imagine a parent, a high-school friend or a (sports) team-mate. Such a person would be willing to read a few pages you have written. At the same time, you want them to find something to carry away. The ideal reaction would be
"Wow! I'd heard of the topic, but never thought it could be so simply explained. Thanks."
For your class mates, you want to convey an intriquing new-to-them idea so clearly, they instinctly ask questions about either the science or a possible application.
From Science, What we don't know: 125 top question
New Times Annual Year in Ideas: 2003 2004 2005 2006 2007
Here are some recent new "physics ideas" in the last few years
One person per topic Note: Produce a sentence outline and discuss it with instructors before writing.
Here are some broad areas. Online references are helpful. But don't forget books in the State's libraries. While interlibrary loans are fast, you still should start early gathering references.
A policy topic is one where someone else, with the same facts but different priorites, could argue the opposite conclusion.
For example: "how GPS-guided missles work" is not a policy topic. "The U.S. should not use GPS-guided missiles because doing so would make the GPS satellites targets vulnerable and the GPS system too valuable for civilian uses" is a policy topic. Likewise, "Planes using GPS-based guidance systems should be required to have a non-GPS-based backup, since the GPS is a military system and GPS-guided missiles may cause the GPS satellites to become targets" is also a policy topic.
As another example, "how the overuse of antibiotics leads to antibiotic resistance" is not a policy topic, but "anitibiotic use should be regulated to reduce antibiotic resistance" or "anitibiotic use should not be regulated because such regulation would be expensive and cost lives" are policy topics.
Also, the reasons for your policy must be science-based. "Animals should not be used for testing cosmetics because it is cruel to the animals" is not a science-based policy. "Animals should not be used for testing cosmetics because the approach is less reliable than computer modelling" is a science-based policy (it is not clear that if is sustainable, but that is your problem).
Are these possible topics?
Antibiotic regulation, animal testing, national missile defense, enforced zero population growth, fission generated electricity, fusion funding, Kyoto emission trading, high-power laser pointers, lie-detector tests as evidence, reprocessing of nuclear waste, food irradiation, regulation of dietary suppliments.
Or these? Restrictions on cloning of human beings (mostly ethics, not science), evolution/creationism (will end up being all policy or all science)?