Effect of buffer gas on an electromagnetically induced transparency in a ladder system using thermal rubidium vapor

We report on the observation of electromagnetically induced transparency in a ladder system in the presence of a buffer gas. In particular, we study the $5S_{1/2}$-$5P_{3/2}-5D_{5/2}$ transition in thermal rubidium vapor with a neon buffer gas at a pressure of 6 Torr. In contrast to the line-narrowing effect of buffer gas on $\Lambda$ systems, we show that the presence of the buffer gas leads to an additional broadening of $(34\pm 5)$ MHz, which suggests a cross section for Rb($5D_{5/2}$)–Ne of ${\sigma }_{\mathrm{k}}^{(D)}=(23\pm 4)\times {10}^{-19}\hspace{0.5em}{\mathrm{m}}^2$. However, in the limit where the coupling Rabi frequency is larger than the collisional dephasing, a strong transparency feature can still be observed.

Figure 1

(a) Schematic of the Rb vapor cell. The cell consists of two separate regions: a region with thickness of 10 mm and a wedge, where the thickness varies between 300 nm and 3 $\mu$m over a region of approximately 2 cm. (b) Schematic of the experimental setup. A reference cell of length 10 cm is used for frequency calibration. A photodiode PD1 (PD2) is used to record the signal from the buffer cell (reference cell). NDF, neutral density filter; PBS, polarizing beam splitter; BPF, band pass filter; and $\lambda /2$, half-wave plate.

Figure 2

(a) Diagram of the level scheme to clarify the symbols used in the text. (b) Transmission spectra as a function of the probe detuning through the 10-mm buffer gas cell, measured at $T=(52\pm 1)$${}^{°}$C without the coupling beam (dashed red line). The power in the probe beam was $P_{\mathrm{p}}=1\hspace{0.5em}\mathrm{\mu }\mathrm{W}$. The zero of the detuning scale is set to the frequency of the ${}^{85}\mathrm{Rb}$ 5$S$${}_{1/2}{F}_{\mathrm{g}}=3\to$ 5$P$${}_{3/2}{F}_{\mathrm{e}}=4$ transition. The left spectral line corresponds to the ${}^{87}\mathrm{Rb}$ 5$S$${}_{1/2}{F}_{\mathrm{g}}=2\to$ 5$P$${}_{3/2}{F}_{\mathrm{e}}=1,2,3$ transition. Switching on the coupling beam (blue line) with a power of $P_{\mathrm{c}}=196$ mW results in the coupling of $F_{\mathrm{e}}$ to the 5$D$${}_{5/2}$ manifold.

Figure 3

Transmission of the ${}^{85}\mathrm{Rb}$ $F_{\mathrm{g}}=3\to F_{\mathrm{e}}=2,3,4$ transitions through the $~780$-nm-thick region of the cell, (a) without and (b) with 6-Torr Ne buffer gas, for coupling laser powers of (top to bottom) 0, 4, 14, 28, 70, 112, 154, and 196 mW. The power of the probe laser was 1 $\mu$W, and the temperature of the cell was ($430\pm 10$) K. An offset in transmission of 0.075 has been added for clarity to all spectra except the topmost.

Figure 4

Width of the EIT feature against coupling laser power for the ${}^{85}\mathrm{Rb}$ 5$S$${}_{1/2}{F}_{\mathrm{g}}=3\to$ 5$P$${}_{3/2}{F}_{\mathrm{e}}=2,3,4\to$ 5$D$${}_{5/2}$ transition without ($\blacksquare$) and with buffer gas ($•$), respectively. The widths were obtained by a fit with a Gaussian function to the data shown in Fig. 3. The thickness of the atomic sample is $~780$ nm. The width shows a linear dependence on the coupling laser power with a slope of $(900\pm 50)$ MHz/W. Adding 6 Torr of Ne buffer gas to the atomic sample results in a EIT linewidth in the low power limit ($P_{\mathrm{c}}\to 0$) of $2\pi \times (46\pm 5)$ MHz. Note that the error bars, given by the standard deviation, increase as the Gaussian fit function becomes inaccurate in the Autler-Townes limit.