Visible and Ultraviolet Plasma Lines of the He–Ne Gas Laser

Abstract

A study of helium–neon laser plasma lines was done using a double grating spectrometer and a helium-neon laser with an emission wavelength of 632.8 nm (15 802 cm⁻¹). The absolute wavenumber, measured to within $\sim$ 0.2 cm⁻¹, and wavelength of each plasma line are presented, along with intensity and shift relative to the main laser line. Several of the measured lines have not been reported in the literature and are observed at shifts between 0 and 1500 cm⁻¹ from the laser line, a spectral region commonly probed by optical Raman scattering experiments. Accounting for the possibility of second-order diffraction permitted many previously unassigned lines to be attributed to known neon electronic transitions with wavelengths in the ultraviolet region of the electromagnetic spectrum.

Graphical abstract

This is a visual representation of the abstract.

Introduction

Despite increasing adoption of solid state lasers, the helium-neon (He–Ne) gas laser is still one of the most commonly used light sources in research and industrial applications. Like all gas lasers, He–Ne lasers often emit so-called plasma lines arising from spontaneous emission in the laser tube in addition to the desired stimulated emission resulting from population inversion. In laser spectroscopy experiments, where precise knowledge of the wavenumbers associated with spectra features of interest are required, these plasma lines can serve as “built-in” calibration standards. It is, therefore, surprising that the literature on plasma emission lines for He–Ne lasers is scant.

Previous work by Loader ¹ appears to be the only in-depth study of He–Ne laser plasma emission lines. In this study, 38 plasma lines were observed over a spectral range of 12 093–15 798 cm⁻¹. The assignments of these lines and quoted wavenumbers, however, are questionable because the same study also includes results for Ar⁺ laser plasma lines which have since come under scrutiny due to the apparent misidentification of several spectral peaks as plasma emission lines and to uncertainty in measured wavenumbers resulting from use of a spectrometer with relatively low resolution. ² There are also differences in He–Ne plasma line wavenumber assignments as reported by Loader and those in a Raman scattering study on multiferroic compounds. ³ In particular, Raman spectra of CuFeO₂ and CuCrO₂ yielded strong plasma lines in the vicinity of 137, 180, and 200 cm⁻¹, the former two differing by $\ge 2$ cm⁻¹ when compared to the previous data presented by Loader, and the latter one being completely absent.

Motivated by the above-noted paucity of literature data and the inconsistencies in that which is available, we present in this article the results of a comprehensive study on plasma lines from a He–Ne laser with an emission wavelength of 632.8 nm. We report plasma lines with shifts of 0–1500 cm⁻¹ from the primary emission line at 15 802.4 cm⁻¹, a commonly probed spectral range for optical Raman scattering studies, and tabulate absolute wavenumber, measured intensity, and wavenumber shift from the primary emission line. This information is also provided for some second-order lines in the ultraviolet spectral region. In addition, the atomic species associated with each transition is identified.

Experimental Details

Figure 1 shows a schematic diagram of the apparatus used to collect the emission spectra. The experiments were done in air at room-temperature with a He–Ne laser emitting at a wavelength of 632.8 nm. The incident power was $\le$ 5 mW, making observation of unwanted peaks due to inelastic light scattering much less likely compared to higher incident power levels. The beam was focused onto one of the three samples using a $f=5$  cm lens. The reflected/scattered light was collected by the same lens (i.e., backscattering configuration) and focused onto the entrance slit of a Spex model 1401 double grating spectrometer using a lens of focal length $f=25$  cm. All three spectrometer slits (entrance, center, and exit) were set to a width of 350 $\mu$ m. Using holographic gratings with 1800 grooves/mm, the estimated spectral resolution is $\sim 3$  cm⁻¹. The light exiting the spectrometer was detected by a low dark count rate ( $\sim 1$  s⁻¹) photomultiplier tube.

Figure 1.

Experimental set-up. M: Front surface mirror, P: prism, L: lens, and S: sample.

Experiments were conducted over shifts of 0–1500 cm⁻¹ relative to the laser wavelength at 632.8  nm with crystalline silicon, aluminum, and superconductor Bi₂Sr₂CaCu₂O ${}_{8.17}$ as scattering sources to ensure that the origin of peaks assigned to plasma lines was indeed emission from the laser tube and not inelastic scattering from a particular material. To ensure proper calibration of the spectrometer before each run, the signal from the primary emission line was first obtained over the range 15 809.8–15 800.5  cm⁻¹ with a grating step size of 0.05  cm⁻¹ by placing several neutral density filters in the beam path to heavily attenuate the incident light. Following this, the filters were removed and spectrum collection was resumed (with no change in scan direction so as to eliminate possible dial backlash issues) at a starting wavenumber of 15 800.5 cm⁻¹ and a step size of either 0.15 or 0.3 cm⁻¹, corresponding to total acquisition times of $\sim 222$ and $\sim 111$ h, respectively.

Results and Discussion

Figure 2 shows spectra collected in air of laser light scattered from each of the three materials. The spectra are very similar to one another, suggesting that our assignments of multiply-observed peaks to plasma emission lines are robust. In order to be classified as a plasma line, it was required that the line be present in spectra of all three materials at the same shift. The peak positions reported in this article are those obtained for c-Si as the scattering material because its spectrum was collected for the longest time with the smallest step size (0.15 cm⁻¹). As shown in the top and bottom panels of Figure 2, the spectral regions between 0–300 cm⁻¹ and 1200–1500 cm⁻¹ contain particularly large number of plasma lines. In contrast, the center panel of Figure 2 shows that relatively few lines are present in the spectral region from 400 to 800 cm⁻¹. Spectral peak parameters were obtained by fitting a Voigt function to each peak using the “Fityk” routine. ⁴ All lines were found to have a width of $\sim 0.3$  cm⁻¹, suggesting a common origin, thereby providing further support for their assignment as plasma emission lines.

Figure 2.

Emission spectra of a Helium-Neon (He–Ne) laser with primary emission wavelength of 632.8 nm obtained with three different scattering sources. Multiple plasma lines are present at the same shifts in all three spectra. Top panel: Crystalline silicon (c-Si); middle panel: aluminum (Al); bottom panel: Bi₂Sr₂CaCu₂O ${}_{8.17}$ (Bi-2212). Insets show zoomed-in spectral regions of particular interest.

Comparison of our spectra to the data presented by Loader reveals multiple plasma peaks that were not observed in the latter study (the criterion for being labelled as a “different” line being that the shift differed by more than $\sim 1$  cm⁻¹ from the closest line in our spectra). Our spectra in Figure 2 also confirm the presence of a presumed plasma line at a shift of $\sim 200$  cm⁻¹ reported in a previous Raman scattering study which we assign to a Ne electronic transition.^3,5

The absence in the literature of multiple lines that appear in our spectra is surprising, especially given that our results indicate that the intensity of many of the these lines is greater than that of some previously observed. ¹ This can be largely reconciled by noting that ultraviolet lines corresponding to known Ne transitions with emission wavelengths of $\sim 300$  nm can satisfy the grating equation for second-order diffraction. ⁵ With these points in mind, we present in Table I the wavelength in air ( ${\lambda }_a$ ) and vacuum ( ${\lambda }_v$ ), absolute wavenumber in air ( ${\nu }_a$ ), and vacuum ( ${\nu }_v$ ), shift relative to the 632.8 nm primary emission line ( $\mathrm{\Delta }{\nu }_a$ ), and the intensity of each of the observed plasma emission lines as shown in Figure 2. We also list the Ne electronic transitions likely responsible for the observed lines. Analogous data presented by Loader are also included for the purposes of comparison. ¹

Table 1.

Wavelength in air ( ${\lambda }_a$ ) and vacuum ( ${\lambda }_v$ ), absolute wavenumber in air ( ${\nu }_a$ ) and vacuum ( ${\nu }_v$ ), wavenumber shift relative to 15802.4 cm⁻¹ in air ( $\mathrm{\Delta }{\nu }_a$ ), and intensity of plasma lines for a He–Ne laser emitting at 632.8 nm. ⁵

Study	${\lambda }_v$	${\nu }_v$	${\lambda }_a$	${\nu }_a$	$\mathrm{\Delta }{\nu }_a$	Intensity	Diff. in ${\nu }_v$	Neon transition
	(nm)	(cm−1)	(nm)	(cm−1)	(cm−1)	(A.U.)	(cm−1)	(nm)
PW (m = 1)	632.991	15798.0	632.816	15802.4	0	–	–	–
	633.187	15793.1	633.012	15797.5	4.9	1421	–	633.08894
	635.318	15740.1	635.142	15744.5	57.9	305	+1.0	635.18532
	636.640	15707.5	636.464	15711.8	90.2	60	–	636.49963
	638.432	15663.4	638.256	15667.7	134.7	199	+1.0	638.29914
	640.251	15618.9	640.074	15623.2	179.2	1611	–	640.10760
	640.361	15616.1	640.184	15620.5	181.9	85	+0.9	640.22480
	641.112	15597.9	640.935	15602.2	200.2	129	–	640.97469
	642.311	15568.8	642.133	15573.1	229.3	36	+0.9	642.17044
	644.628	15512.8	644.450	15517.1	285.3	111	+0.5	644.47118
	650.823	15365.1	650.643	15369.4	433.0	103	+0.2	650.65277
	653.485	15302.6	653.304	15306.8	495.6	30	−0.4	653.28824
	660.113	15148.9	659.931	15153.1	649.3	49	−0.8	659.89528
	660.515	15139.7	660.332	15143.9	658.5	22	–	660.29007
	666.913	14994.5	666.729	14998.6	803.8	22	−0.8	666.68920
	668.072	14968.5	667.887	14972.6	829.8	59	−1.3	667.82766
	671.940	14882.2	671.754	14886.4	916.0	31	−1.2	671.70430
	693.169	14426.5	692.977	14430.5	1371.9	428	−0.6	692.94672
PW (m = 2)	337.117	29663.3	337.019	29671.9	966.4	133	–	336.99069
	341.932	29245.6	341.833	29254.1	1175.4	541	–	341.80052
	342.529	29194.6	342.430	29203.0	1200.9	37	–	342.39120
	344.918	28992.4	344.818	29000.8	1302.0	766	–	344.77022
	345.221	28967.0	345.121	28975.3	1314.7	148	–	345.07641
	345.568	28937.9	345.468	28946.2	1329.3	549	–	345.47720
	346.203	28884.8	346.103	28893.1	1355.8	345	–	346.05235
	346.813	28834.0	346.712	28842.4	1381.2	626	–	346.65781
	347.409	28784.5	347.308	28792.9	1405.9	1438	–	347.25706
Loader 1		15798.002	632.81646
		15782.381	633.44279
		15739.064	635.18618
		15662.306	638.29914
		15615.202	640.22460
		15567.871	642.17108
		15512.310	644.47118
		15364.935	650.65279
		15302.951	653.28824
		15149.735	659.89529
		15028.714	665.20925
		14995.342	666.68967
		14969.790	667.82764
		14883.395	671.70428
		14427.144	692.94672

$m$ :order of diffraction, PW: present work. The refractive index of air used to convert ${\lambda }_a$ to ${\lambda }_v$ was 1.0002765 for $m=1$ and 1.00029 for $m=2$ . ⁶ Tabulated data from the present work has a wavelength uncertainty of $\pm 0.008$   nm, wavenumber uncertainty of $\pm 0.2$   cm⁻¹, and intensity uncertainty of $\pm 10\mathrm{\%}$ , all of which were obtained by collecting multiple spectra over a specific spectral range and calculating the standard deviation from resulting best-fit peak parameters. Known neon transitions in air are listed in the right-most column.

Table I reveals some notable differences between the current and previously published results. Firstly, several of our measured plasma line wavenumbers differ from those of Loader’s work by $\gtrsim 1$   cm⁻¹ (see second column from the right in Table I). Secondly, plasma peaks were previously reported near shifts of 15.63 and 769.51 cm⁻¹. ¹ We also observe a feature at $\sim 15$ cm⁻¹, but no peak is present in our spectra at the latter wavenumber shift. While not well-resolved, the general shape (broad and weak) of the feature at $\sim 15$  cm⁻¹ does not appear to be characteristic of a typical plasma line (see top inset in Figure 2). We note, however, that there is a known Ne transition at ${\lambda }_a=633.44276$ nm corresponding to a shift in air of $\sim 15.7$  cm⁻¹ from the primary emission line and, therefore, it is possible that the feature at $\sim 15$  cm⁻¹ is in fact a plasma line. ⁵ Furthermore, with regard to the peak reported at a shift of $\sim$ 770  m⁻¹, although absent in our spectra, it also cannot be discounted as a possible plasma line because there are two known Ne transitions close to this relative shift. ⁵ These transitions have an emission wavelength of ${\lambda }_a=665.20925$  nm (first-order diffraction) and ${\lambda }_a=332.715344$  nm (second-order diffraction), resulting in relative shifts of $\mathrm{\Delta }{\nu }_a=769.53628$  cm⁻¹ and $\mathrm{\Delta }{\nu }_a=774.53883$  cm⁻¹, respectively.

Conclusion

Plasma lines from a He–Ne laser source were investigated by probing a spectral region 0–1500 cm⁻¹ from the primary emission wavelength of 632.8 nm ( ${\nu }_a=15802.4$  cm⁻¹). From the collected spectra, multiple plasma lines not observed before with a grating spectrometer were identified as electronic transitions of Ne. The shifts of several of the lines are also consistent with second-order diffraction in the UV region of the electromagnetic spectrum. This could be easily validated by placing a long-pass filter in front of the entrance pinhole of the spectrometer.

This work complements published He–Ne laser emission line compilations and refines previously measured plasma peak shifts through use of a spectrometer with higher resolution. With an estimated uncertainty of $\sim$ 0.2 cm⁻¹, the data reported in the present work constitute the most complete and accurate compilation of He-Ne emission line wavelengths to date and should prove useful to those performing spectroscopy experiments that employ a He–Ne laser as a calibration or incident light source in the visible and ultraviolet spectral regions.