1: \begin{abstract}   

2: 

3: We calculate the complete one-loop effective potential

4: for $SU(2)$ gauge bosons at temperature $T$ as a

5: function of two variables:

6: $\phi $, the angle associated with a non-trivial Polyakov loop,

7: and $H$, a constant background chromomagnetic field.

8: These two variables are indicators for confinement and

9: scale symmetry breaking, respectively.

10: Using techniques broadly applicable to finite temperature field theories,

11: we develop both low and high temperature expansions.

12: At low temperatures,

13: the real part of the effective potential $V_R$

14: indicates a rich phase structure,

15: with a discontinuous

16: alternation between confined $(\phi=\pi)$

17: and deconfined phases $(\phi=0)$. 

18: The background field

19: $H$ moves slowly upward from its zero-temperature value

20: as $T$ increases,

21: in such a way that $\sqrt{gH}/\pi T$ is approximately

22: an integer.

23: This behavior stops at $T_c = 0.722(1) \mu_0$, where $\mu_0$

24: is a zero-temperature renormalization group invariant scale;

25: beyond this temperature, the deconfined phase is always preferred.

26: At high temperatures, where perturbation theory should be

27: reliable as a consequence of asymptotic freedom,

28: the deconfined phase  $(\phi=0)$ is always preferred, and

29: $\sqrt{gH}$ is of order $g^2(T)T$.

30: The imaginary part of the effective potential $V_I$,

31: which originates in a tachyonic mode associated with the lowest

32: Landau level,

33: is non-zero

34: at the global minimum of $V_R$ for all temperatures.

35: A non-perturbative

36: magnetic screening mass of the form

37: $M_m = c g^2(T) T$ with a sufficiently large coefficient $c$

38: removes this instability at high temperature,

39: leading to a

40: stable high-temperature phase with

41: $\phi = 0$ and $H=0$, characteristic of a weakly-interacting gas of

42: gauge particles. 

43: The value of $M_m$ obtained is comparable with lattice estimates.

44: 

45: 

46: \end{abstract}

47: