You've got the start of a good concept here. It doesn't look like much, but everyone knows programmer art is rarely presentable enough for a release version. The scientist looks pretty good, though.
The controls are a little unresponsive. The grid-based movement makes sense for this kind of game, but the scientist's start-and-stop motion makes me feel like I'm pushing a heavy boulder around the level. There's also no reason he couldn't move faster; puzzle games by their nature tend to be fairly slow-paced, but once I know where I want the scientist to go I want him to get there quickly, if you see what I mean. Also I noticed that standing next to the buttons wasn't sufficient to push them and I instead had to run into them in order for the game to know which button I was pushing, which to me raised the question of why I can't just run into the button to press it rather than having to press Z.
You have a pretty good grasp of level design in my opinion. I didn't get stuck until several levels in when I entered a room with a purple teleporter but no means of actually turning it on, which is a shame because at that point I had been under the impression you'd gone to great lengths to avoid creating a situation where the player could trap themselves like that. Even dying would have been preferable to a soft lock, but whatever, it's forgivable for an early demo.
Spice up the levels with some scenery. It will provide some idea of context, create a sense of the scientist's location and ideally build an atmosphere like this is a place where people work and possibly where something has gone down that the scientist is trying to escape, maybe even build a real narrative or a story to add some motivation to the puzzles.
This is a pretty basic game, but I can see that you've got some good ideas here so I'm gonna give you a solid 3 stars. Lick of paint, buff out some places, and you'll have yourself a pretty decent amateur game project.