Xiaochao Zheng, xiaochao@jlab.org

For the polarized 3He target group @ Jlab Hall-A

June 25, 2001

• What is Masing Effect?
Here is a quantitative description of the masing effect. The basic idea is the following: Consider a polarized cell in a holding magnetic field along the z direction. The cell is placed inside (or near) a coil which is part of an LC circuit. Assume that the magnetization has a small transverse component. The spins are precessing around the magnetic field. Their precession induces a voltage in the coil. The voltage induced in the coil causes a current to flow through it. This current produces a magnetic field transverse to the holding field. Under certain conditions this induced field can cause the spins to tip away from the z axis and increase the voltage induced in the coil. That in turn increases the transverse field an dcauses a runaway situation. The longitudinal polarization of the spin will decrease while certain conditions are met. There after it will remain constant at the so-called masing threshold, which can be determined by the steady solution of spin equations.

• What are the typical phenomena of Masing Effect?
• Masing can occur only if the spins are polarized in the high energy Zeeman state, so they can dump ther energy into the coil.
• Masing is characterized by a threshold, below which the effect is negligible (the polarization is stable) and above which the polarization is unstable and will spontaneously develope a large transverse component.
• The value of the threshold depends on the coupling to the coil, which is dependent on the difference between the resonance frequency of the coil and the Larmor frequency of the spins.
• The threshold depends on the transverse nuclear relaxation time T2. It becomes larger when T2 is short, for example, because of a large field gradient.

• How to reduce Masing Effect?
Masing is a major effect which harms the target polarization and we certainly do not want it to happen during the experiment. There are a few things which will help to reduce masing or let it completely gone.

As was mentioned in the last section, "the threshold depends on the transverse nuclear relaxation time T2. It becomes larger when T2 is short, for example, because of a large field gradient." In other words, an additional gradient field will destroy the certain conditions which form the masing effect.

Based on this idea a magnetic field gradient has been applied. This can be done by attaching a gradient field coil to the main holding field Helmholtz coils. It, indeed, caused the polarization to rise sharply in both EEL lab tests and during the real runs of E99-117. It should be pointed out that the more usual relaxation due to field gradients would tend to decrease the polarization so one needs to minimize field gradient as far as masing is gone.

• Changing main holding field
A smaller holding field can decrease the coil coupling as the Larmor frequency is moved further from the coil resonance. So even in the absence of a field gradient the polarizatin tends to rise at smaller holding field. This has been proved during target tests at SLAC, where the holding ifled was decreased from 19 Gauss to 9 Gauss.
But during our experiment E99-117/E97-103 we did not try this. First of all, no test has been done at EEL target lab so we do not have any data/experience on this issue. Next, both NMR and EPR polarimetries will be affected if the holding field changes. So lowering main holding field will be our last choice.

• What does it mean by "NMR/EPR triggers masing effect?"

NMR and EPR are the two basic polarimetries we have been using for polarized 3He target. Both of them need to flip 3He spins to perform the measurement. Usually the "certain conditions" for masing effect to happen are quite subtle and difficult to satisfy. AFP spin flip during NMR/EPR provides a perturbation which help to form these certain conditions. So typically masing effect will not happen until an NMR/EPR is performed, when masing will happen spontaneously. In other words, if no polarimetry is performed, then target polarization will stay high. But then we will not be able to know the polarizaiton.

• EEL lab test?
During EEL target lab tests in Feburary, we attached a 20-turns gradient coil to the small Helmholtz coil (which was orientated at an angle of 19 degrees w.r.t. beamline). Three series of tests were done with cell "GORE" at gradient coil current 0A, 5A and 7A. EPR measurement shows:
 (1)with gradient coil current 0 A, target polarization (in pumping chamber) dropped from 31% to 23% in 25 minutes, stayed there for 3 hours, then slowly went back up. See Figure 1. This shows that masing threshold without field gradient is 23%.
 (2) with gradient coil current 5 A, polarization dropped from 43% to 39% in 20 minutes, stayed there for 3 hours, then slowly went back up. See Figure 2. This indicates that masing threshold with field gradient 5 A is 39%.
 (3) with gradient coil current 7 A, polarization dropped linearly as EPRs were performed, which can be explained by normal AFP loss due to EPR measurements. See Figure 3. Polarization was continuously dropping because only two lasers (compared to thress) were used during this test.

• Comments: These 3 tests show typical masing effect on GORE. By applying a gradient current of 7 Amp (corresponds to field gradient ??? mG/cm) masing was gone.

• Notes: By NMR we also did a few tests to study masing effect. But the masing we observed on GORE probablly starts from pumping chamber then slowly transfers to target chamber. Since EPR measures the polarization inside pumping chamber while NMR is for target chamber, so EPR is more sensitive to masing effect here.

• Masing during E99-117 - Period I.

Gore was used at the beginning of E99-117 in June, 2001. During the first week of running no masing effect was observed and Gore reached its maximum 46% after 3 days of pumping. But after the first maintainance period (June 11), masing suddenly happened. See Figure 4:

Target polarization dropped from 43% to 30% whenever an NMR(red)/EPR(blue) was performed (June 13 and June 14). This cannot be explained by normal AFP losses or depolarization due to electron beam. It was unfeasible to do masing tests during data taking but based on theory/experience we believe that masing was back with a threshold of roughly 30%.
At 5pm, June 14th, a gradient coil with current 5 Amp was applied, which was attached to small (now longitudinal) Helmholtz coil. EPR measurements show that masing threshold has been improvedd to 37%. Then gradient current was increased to 7 Amp around June 14th 11pm. EPR measurements after that show a steady polarization of 38%.

• Masing during E99-117 - Period II.

Masing was not completely gone at a gradient current of 7 Amp. In principle we should increase the current. But new problems arised. First is that NMR signal was distorted due to large field gradient which caused NMR signal be difficult to fit. This eventually will introduce a large error to NMR polarimetry. Next, 7 Amp is the highest current of the power supply which was using. Third of all, it has been noticed that field gradient in longitudinal direction is not successful in reducing masing. (Imagine the field gradient applied to small coil during EEL lab tests has a large transverse component of 7 Amp * cos(19 degrees)).

It has been expected that a second gradient coil at transverse direction will be more efficient in fighting against masing. Based on this idea, a 10 turns coil has been attached to transverse Helmholtz coil on June 19th. After a few days of testing now we are staying at a field gradient setting of transverse 6.00 Amp and longitudinal 2.00 Amp. The maximum (in beam) polarization now is 39% at holding field 0 degree and 37% at 180 degree.

• Why is masing different at 0 and 180 holding field directions?

It should be noticed that the field gradients usually have two components, one from the imperfections of the Helmholtz coils, which is proportional to the magnetic field, and the other from the ambient field gradients. When the holding field is reversed, the two components add or subtract in different ways, which can lead to a slightly different value of T2 and change the masing threshold.

• Why do you switch the polarity of gradient field coil back and forth?

Based on the same idea as above, the effect gradient field acting on masing is different at different gradient field directions and we were trying to optimize the settings by switching the polarity of gradient coil.

• Is MASING amazing?

Masing is described by non-linear equations. In principle it is possible to solve all equations and calculate masing threshold. But in practice there are various parameters unknown/not understood contributing to masing, which make the numerical estimates of masing threshold difficult and unsuccessful.

So far we have a lot of qualitative evidence for the masing effect, in fact we cannot think of any other effect which could explain even a fraction of the spin behavior explained by masing. But we have never been successful in calculating masing threshold and making comparison with data.

In one word, we know masing is there when it happens, but we cannot see its face clearly.