I want to know if its possible to create only a magnetic field,which i can direct to a certain chosen object?
Yes but please look at the paper by Arbab at the arxiv link http://arxiv.org/abs/1301.4719 Best wihes.
Your question is best taken in two parts: (1) Are magnetic (or mostly-magnetic) waves possible, and (2) If they do exist, would it be possible to use such waves to implement the science fiction concept of "tractor beams" -- that is, as a way to exert a spatially focused, selective pull (or push) on a relatively distant object? So, here goes:
(1) Are mostly-magnetic waves possible?
While it is true that you can never generate a purely magnetic wave, you can arrange things so that at a first approximation it looks like you are simply modulating a magnetic field without dissipating much energy as conventional radio waves. If you want a simple and highly accessible example, place a ball or cylinder magnet on top of a table and wiggle a strong magnet beneath the table. Your top magnet will move in a cyclic fashion, because it's being moved by a repeating wave of magnetic force from below the table.
You could describe that force using a full electromagnetic wave description, but to be blunt about it, why bother? Describing what you have just done as a "mostly magnetic" wave works quite well, since the behavior of the top magnet depends almost entirely on changes in the local intensity of the magnetic field generated from beneath the table.
If that example makes the idea of magnetic waves sound too trivial, it's worth pointing out that no less a figure than Nikola Tesla once proposed such waves as a way to build communications systems that would compete with Marconi's radio systems. That's rather ironic, incidentally, since by most accounts Marconi made extensive unauthorized use of Tesla's patents to build his radio systems.
However, your question has more than just historical significance. Recently, Lockheed Martin resurrected the idea of mostly-magnetic waves to create a system that allows trapped miners to communicate with the surface. Such communication is flatly impossible using conventional radio frequency bands, since the electric component of such waves is very quickly absorbed by the earth. Mostly-magnetic waves in contrast are scarcely absorbed by rock, earth, or water, and therefore do hugely better at penetrating upwards through the earth to reach the surface.
So, not only is your question a good one, it has a noble history (Tesla) and has recently resulted in an important invention. Magnetic waves have created a renaissance in communications under very difficult circumstances, resulting in devices that no doubt will be saving lives around the globe for years to come.
[And yes, before someone asks: By symmetry you can also create "mostly electric" waves by moving either isolated strong electrostatic charges or (more controllable) large-scale static dipoles back and forth at relatively slow speeds. The practical example is moving a charged balloon towards and away from your hair, making your hair "wave" in the resulting mostly-electric wave pattern. Electric waves likely would be less practical than their magnetic equivalents, since mostly-electric waves will be strongly absorbed by most forms of ordinary matter. But the bottom line is that through symmetry arguments alone, both "mostly magnetic" and "mostly electric" have to be equally plausible. I am not aware of any applications, or even any references, to mostly-electric waves, by the way. I'm writing this part of my answer based only on symmetry arguments.]
(2) Can mostly-magnetic waves be used to create a spatially selective pull or push on a distant object -- that is, a tractor beam?
This is actually the harder question. I don't think so, but I would have to look a lot more closely at the options possible before stating so definitively.
Electromagnetic waves in general create both magnetic and electric dipoles, and those can in fact be directed so that they occur at a considerable distance away. A laser is the classic example: It creates a rapidly alternating electric dipole -- one that changes direction at the frequency of the light -- at potentially very remote locations. So, you can most definitely use electromagnetics to create transient "pushes" of electric and magnetic forces at distance sites.
Recall how you can move magnets slowly under a table to create a movement on top as an example of a mostly-magnetic wave? Your question about focusing magnetic force at a distance amounts to roughly to this: Is it possible to create a distant, spatially focused magnetic dipole at a sufficiently low frequency (and sufficiently high power) that, on a human time scale, it results in a ferrous or magnetic object being moved in some desired direction?
Oddly, the safest answer to this question probably is "yes" (surprised?). That's because it's basically a scale up question, and electromagnetic phenomena scale up very well in principle, even if not in practice. That is, if you can use a laser to create a very brief but well-directed magnetic dipole at a remote location, you should also with the right set up be able to create a much longer-lasing magnetic dipole by using much lower frequencies (and I do mean much lower, e.g. a tenth of a cycle per second). The problem is that given the very low efficiency of such antennas, the "laser" for that puppy would likely be some kind of antenna occupying most of Nebraska, for a very rough guess (anyone?) The side effects of the equally strong electric vectors would be... interesting.
However, that's using conventional electromagnetic waves. Since mostly-magnetic waves certainly do exist and are in use, the question changes form a bit to this: Can mostly-magnetic waves be configured so that as with laser radiation, they mostly self-cancel in all directions but a selected one? That is, for example, could some clever array of moving neodymium magnets create magnetic fields that would reinforce in some directions and cancel in most or all other directions, so that their slow, human-time (recall the table top example) pull could be projected a greater distance than is possible with isolated magnets, whose fields fall off very quickly with distance.
I do not know the answer to this. My temptation is to say "no," but I know that's mostly an emotional reaction. That is, I can't quote a precise reason why that would necessarily be prevented by electromagnetics. After all, by scale-up analogy it clearly is possible in principle (even if not in practice) to create a strong magnetic dipole lasting seconds at selected distant locations. I don't see any obvious short-cut invocation of some principle that would absolutely forbid self-cancellation and self-reinforcement in mostly-magnetic waves cases, one that would keep it from enabling projection of a magnetic dipole further out than is possible with isolated magnets. Someone? Anyone?
If you take a solenoid with a low frequency alternate current, the corresponding magnetic field will be accompanied with a rather weak electric field due to low frequency. But this magnetic field decays with distance very quickly, it is variable, but it does not propagate, and it is called a near field. The propagating part of the total electromagnetic filed will be equally weak both in magnetic and electric parts due to the low frequency.