Light-induced atom desorption for cesium loading of a magneto-optical trap: Analysis and experimental investigations

With the help of light-emitting diodes at the center wavelength of 470 nm, we demonstrate that light-induced atom desorption (LIAD) can be used for flexibly controlling the loading of magneto-optical traps (MOT) of cesium atoms. Under an ultralow background pressure in a quartz cell without any wall coating, we show that low intensity blue light can be used to control the loading rates (from 200 to 4000 atoms/s) and the number of cesium atoms (on the order of ${10}^4$ in our experiment) in a MOT without the use of dispensers or any secondary atom source. A theoretical model based on an atom loading rate equation is built which can simulate the magneto-optical trap loading process with LIAD. The theoretical results are in agreement with experimental data. Some important parameters of the vacuum system are determined accordingly. The decay time of the vacuum pressure is about 70 ms, which is much shorter than the usual vacuum system for experiments on atomic ensembles. It is very difficult to measure such short times by the traditional fluorescence detection of cold atoms due to the slow loading process. The change of desorption rate (on the order of ${10}^{13}\text{atoms}/\text{s}$) for the different desorption intensities is also determined based on the background pressure caused by noncesium atoms in the cell. According to the experimental data and theoretical calculations, we obtain a partial pressure of about $1.4\times {10}^{-10}\text{Torr}$ for the untrapped cesium atoms and about $6.4\times {10}^{-10}\text{Torr}$ for the noncesium atoms at the moment the desorption light is turned off. The system is almost an ideal vacuum system for transferring atoms into another region for further experiments.

Figure 1

(Color online) Model of the MOT loading process: two dynamics processes are taken into account: the LIAD process and the MOT loading process. The density of the cesium in the cell $n\left(t\right)$ is determined by the absorption coefficient $\left(A\right)$ and the desorption rate $\left(D\right)$. $1/A$ describes the absorption rate by the surfaces of the glass chamber and other objects inside the chamber; $D$ stands for the atom number desorbed per second. $N\left(t\right)$ is the atom number of the MOT.

Figure 2

(Color online) (a) Time evolution of the atom number with different desorption rates from $1.0\times {10}^{13}$ to $5.0\times {10}^{15}\text{atoms}/\text{s}$. (b) The maximum atom number (solid red line) and the loading rate (dashed green line) versus the desorption rate $D$. The maximum atom number increases very quickly at the beginning when the desorption rate increases but it is saturating at the higher desorption rates. Parameters: $T=300\text{K}$, $m=2.21\times {10}^{-25}\text{kg}$, $V=1.35\times {10}^{-14}{\text{m}}^3$, $\sigma =2\times {10}^{-17}{\text{m}}^2$, $A=14{\text{s}}^{-1}$, and $c=1/11.2{\text{s}}^{-1}$.

Figure 3

(Color online) (a) Time evolution of the atom number versus various adsorption coefficient $A$. (b) The maximum atom number (solid red line) and the loading rate (dashed green line) versus the adsorption coefficient $A$. The large adsorption coefficient corresponds to the lower maximum atom number and lower loading rate. Here $D=15.5\times {10}^{13}\text{atoms}/\text{s}$ and all other parameters are the same as Fig. 2.

Figure 4

(Color online) (a) Time evolution of the atom number versus the coefficient $c$. (b) The maximum atom number (solid red line) and the loading rate (dashed green line) versus the coefficient $c$. The loading rate has almost no change with the increasing of loss rate from the noncesium atoms in the background, while the maximum atom number in the MOT is strongly affected by the background gas. Here $D=15.5\times {10}^{13}\text{atoms}/\text{s}$, $A=14{\text{s}}^{-1}$ and all the other parameters are the same as Fig. 2.

Figure 5

(Color online) Experimental system. The MOT is performed in a transparent quartz cell. The cesium dispenser is positioned the back of the cell with a distance of about 70 mm from the MOT. The desorption light source made by seven bunched blue LEDs is located about 70 mm above the MOT. The fluorescence emitted from the cold atoms in the MOT is collected by a specially designed objective with numerical aperture $\text{NA}=0.29$ and is detected by a high-speed, low-noise CCD camera. There is a filter at center wavelength of 852 nm in front the CCD. An aspheric lens mounted on the oxygen-free copper is used to focus the dipole trap beam for optical microtrap of the single atoms. All of these are indicated in the figure.

Figure 6

(Color online) Time evolution of the atom number with different desorption intensity from 0 to $1.76\text{mW}/{\text{cm}}^2$. The colored solid triangles are the experimental data and the solid black lines are the theoretical fittings according to Eq. (8). The theoretical fittings are good in agreement with the experimental results. From the fittings we obtain the adsorption coefficient $A=\left(14.0\pm 0.1\right){\text{s}}^{-1}$. The loss rate caused by the noncesium atoms in the background is also determined as $c=\left(1/11.2\pm 1/113.6\right){\text{s}}^{-1}$. We also obtain the corresponding desorption rates, which varies from $D=6.2\times {10}^{12}\text{atoms}/\text{s}$ to the maximum of $1.6\times {10}^{14}\text{atoms}/\text{s}$ for each intensity of blue light.

Figure 7

(Color online) The desorption rate (red solid squares) versus the desorption intensity. The solid blue line is the linear fitting. The desorption rates increase linearly with the desorption light at relatively low intensities in our system.

Figure 8

(Color online) (a) Maximum atom number in the MOT versus the desorption intensity. The red solid squares are the experimental data while the solid green line is theoretical fitting according to Eq. (8). The inset shows the zoom of the experiment results and the fitting. The theoretical fittings are good in agreement with the experimental results with the obtained parameters: $A=14.0{\text{s}}^{-1}$ and $c=1/11.2{\text{s}}^{-1}$. (b) The loading rates versus the desorption intensity. The blue solid triangles are the experimental data while the solid red line is the theoretical fitting according to Eq. (8). For the maximum intensity of $1.76\text{mW}/{\text{cm}}^2$, we have achieved the maximum loading rate of about $4.0\times {10}^3\text{atoms}/\text{s}$ in the experiment.

Figure 9

(Color online) The decay time versus the desorption time (red solid squares) and the desorption intensity (green solid triangles). The decay time is obviously independent on the desorption time and the desorption intensity.