2.1 Electronics layout and General Procedure

In order to test the ZDC gain under different high voltages, we constructed an electronics layout as shown in Fig. 3.

Fig. 3.Full-screen View

Electronics layout of the ZDC gain test under different HVs.

Muons from cosmic rays hit sequentially the five modules: the east SMD, the west SMD, the top trigger scintillator, the ZDC module, and the bottom trigger scintillator. Since there are background particles everywhere at anytime, it is essential to use the logic "and" result of the top and bottom scintillators to select events triggered by cosmic ray muons. When a muon hits the ZDC after traversing the top scintillator, it will produce Cherenkov light. The light is guided by the optical fibers and generates a signal in the ZDC. The muon will then hit the bottom scintillator and also generate a signal. The coincidence of the top and bottom scintillator signals is used as the system trigger and provides a gate to the ADC.

2.2 Result

The expected relation of PMT module gain with high voltage should be a power law: gain = a×HV^b, where a and b are parameters related to the PMT module and environment, and the gain is in unit of pC and voltage is in unit of kV. Ideally b is related only to the PMT, so as long as we are using PMT modules of the same standard, b should not vary much. The gain vs. HV plots of different ZDCs, and the fitting results, are shown in Fig. 4. The fit parameters are listed in Table 1.

Fig. 4.Full-screen View

The ZDC gains test under different HVs. The ZDC 2-1, 2-2 and 2-3 are east modules, and other three are west modules.

2.3 Conclusion

As we can see from the result shown in Fig. 4, different fit lines are nearly parallel to each other, which is reasonable since the slope b is a PMT-related parameter and it should have nearly the same value. From the relation between the gain and high voltage, we can get separate values of high voltages necessary for different ZDCs to have the same gain.

3.1 Tower gains and high voltage applied

When a neutron flies through three ZDC modules, the gain in each tower decreases from the first tower to the last one due to the energy loss in ZDC material. Ideally, the ratio of ZDC tower gains should roughly be 6:3:1. However, it is difficult to tune the high voltage perfectly in practice as the gains will be affected by many factors like beam conditions, shower leakage, etc. So in this article, we consider the HV setting to be good if the ratios are not far from the ideal value. With the runs listed in Table 2, we will show the ratios with different beam types in Figs. 5, 6 and 7.

Fig. 5.Full-screen View

The tower gains distribution of run 12130083 (Au+Au @ 200 GeV). The panel (a), (b), (e), and (f) are east towers; panel (c), (d), (g), and (h) are west towers. The upper four panels show the gains distribution between tower1/2 and tower 2/3. The mean gain ratios vs. ADC are then shown in the four lower panels. The red straight line is a linear fit; the slope indicates the gain ratio between two towers.

Fig. 6.Full-screen View

The tower gains distribution of run 15122044 (Au+Au @ 200 GeV).

Fig. 7.Full-screen View

The tower gains distribution of run 15186001 (³He+Au @ 200 GeV). All four panels are for east side.

For Au+Au runs at $\sqrt{s_{{}_{\text{NN}}}}$=200 GeV in the year 2011, as shown in Fig. 5, the gain ratios between tower 1&2 and 2&3 for both sides are shown in four upper panels. The mean gain ratios vs. ADC are then shown in the four lower panels. The slope of a linear fit will indicate the ratio between towers. The uncertainties of the slopes are always on the order of 10^-3, so they are not shown. The fitting range varies from plot to plot, because, in the low-x range, the small signal may introduce large errors, and in the high-x range, the shower leakage and other effects will break the linear relation. Figure 5 shows that, on east side, the slope for tower2/tower1 is 0.84 and for tower3/tower2 is 0.36. For west side, the slopes are 0.84 and 0.39 for tower2/tower1 and tower3/tower2, respectively.

For Au+Au runs at $\sqrt{s_{{}_{\text{NN}}}}$=15 GeV in the year 2014, the neutrons emitted from the peripheral collisions will not carry much energy at this low energy. So it’s not possible to obtain the slopes as we did for the run above due to the small gains on towers. We counted the raw gains for each tower and calculated the ratio directly between towers: the ratio is 6.38:1.61:1 on the east, and is 7.35:2.37:1 on the west, where we take the gain of third tower as unit of 1.

For Au+Au runs at $\sqrt{s_{{}_{\text{NN}}}}$=200 GeV in the year 2014, we still plot the gains and the slopes between towers as shown in Fig. 6. In the figure, the slopes for tower2/tower1 and tower3/tower2 are 0.61 and 0.33 for east side, and for west side they are 0.61 and 0.28.

We also have ³He+Au runs at $\sqrt{s_{{}_{\text{NN}}}}$=200 GeV in the year 2014. In this run, we have asymmetric beam types, which is quite different from runs above. After collisions, the neutral residual of the gold ions will hit the east ZDC modules, while the residual of ${}^{\text{3}}\text{He}$ will hit the west modules. The ZDC gains are quite different from east to west. The gold ions contain more neutrons, so the east ZDC gains are larger than west. In Fig. 7 we show the ZDC gains and the slopes, the west side is not shown due to the low ZDC gains. From the figure we can see that the slope of tower2/tower1 is 0.70, while that of tower3/tower2 is 0.30.

For p+Au runs at $\sqrt{s_{{}_{\text{NN}}}}$=200 GeV in the year 2015, the beam types are also asymmetric. From the raw gains, the ratio between towers are 6.88:3.10:1 on the east side and 9.80:2.75:1 on the west side.

As mentioned above, it is quite difficult to tune the high voltages to get ideal gain ratios. In this article, by fitting, or by direct use of raw gains, we evaluate the HV settings by calculating the gain ratios of towers. Considering that the tower gains are affected by many factors, we think those HV settings are acceptable.

3.2 Energy resolution

The single neutron peak from peripheral collisions is used to determine the energy resolution of the ZDC. The energy resolution of ZDC modules has been simulated and tested in a test beam before. In those tests, the resolution is around E/E=20% at E_n=100 GeV [1]. In this section, we will show energy resolution of ZDC in different beam types and energies since 2011. We require a very low Time of Flight detector multiplicity (less than 2) in order to select the peripheral collisions. We also require the TDC value to be within [500, 2500] in order to reduce the background.

For Au+Au runs at $\sqrt{s_{{}_{\text{NN}}}}$=200 GeV in 2011, the single neutron peak is shown in Fig. 8. In this figure, the panel (a) is the single neutron peak for the east side ZDC and panel (b) is for the west side ZDC. A Gaussian function was used to fit the peak and the fitting parameters are shown also in the plot. For both east and west side ZDCs, the E/E=26%.

Fig. 8.Full-screen View

Single neutron peak from run 12130083 (AuAu@200 GeV). Panel (a) is from the east side, and panel (b) from the west side. The black lines in both panels are gaussian fittings. The fitting parameters are also shown in the panels.

For Au+Au runs at $\sqrt{s_{{}_{\text{NN}}}}$=15 GeV in 2014, the single neutron peak can hardly be extracted due to the low beam energy.

As shown in Fig. 9, we can see that on east side, E/E=29%, and on west side, E/E=27%. For Au+Au runs at $\sqrt{s_{{}_{\text{NN}}}}$=200 GeV in 2014, this is consistent with results from the year 2011 run and the early beam test.

Fig. 9.Full-screen View

Single neutron peak from run 15122044 (Au+Au @ 200 GeV).

In ³He+Au runs at $\sqrt{s_{{}_{\text{NN}}}}$=200 GeV in the year 2014,${}^{\text{3}}\text{He}$ is formed by one neutron and two protons, so it’s rare to have multiple neutrons hit the west ZDC modules in one collision. This is why we can see a relatively clean single neutron peak on the west side (Fig. 10b). On the east, the peak is not as well as the east one due to high background (Fig. 10a). From the Gaussian fit, on the west, E/E=40%. This E/E is much larger than the same year and the year 2011 Au+Au collisions results. This may be related to the larger beam-beam crossing angle in ³He+Au runs than in Au+Au runs.

Fig. 10.Full-screen View

Single neutron peak from run 15186001 (³He+Au @ 200 GeV).

In p+Au runs at $\sqrt{s_{{}_{\text{NN}}}}$=200 GeV in the year 2015, the proton contains no neutron, so it’s rare to have neutrons hit on the west side of ZDC modules. And the gold side, like the ³He+Au runs, has large background. The single neutron peak cannot be seen on either side.

3.3 Conclusion

We report the beam test and physics performance of the new installed ZDC in STAR experiment since 2011. With the detail runs listed in Table 2, the ZDC gains between towers were obtained and the ratios between towers were calculated. The ratios in these runs are close to expectation, which indicate that the high voltage applied on these modules are acceptable. We are further studying the single neutron peak for those runs. The energy resolutions are reasonably good although they are a little larger than the early simulation and beam test (from 20% to 27%) at the year 2000 [1]. We also observed that the energy resolution in ³He + Au collisions is not as good as the result in Au+Au collisions, which may be due to larger crossing angle in ³ He (or p) + Au collisions. Our study provides good reference for detector system built-up in the future CSR external target experiment at IMP-CAS facility [9-12].