Classification of Emission Lines of the Group IllB Elements Aluminium Gallium and Indium Excited by Grimm Glow Discharge Plasmas Using Several Different Plasma Gases KAZUAKI WAGATSUMA Institute for Materials Research Tohoku University Katahira 2-! - 1 Sendai 980 Japan In order to clarify the spectroscopic characteristics of the Group IIIB elements in Grimm glow discharge plasmas the relative intensities of their emission lines in the wavelength range 160-800 nm were investigated and tabulated when using different plasma gases oiz. argon neon argowhelium and neowhelium. It was characteristic of the glow discharge light source that the emission spectra observed were strongly dependent on the type of plasma gas employed. Extremely intense ionic lines were observed when neon rather than argon was employed as the plasma gas.Furthermore different groups of ionic lines could be observed only in the helium- containing plasmas. The excitation of these ionic lines caii be principally attributed to charge-transfer collisions between the analyte atoms and the plasma gas ions. Keywords Grimm glow discharge plasma; Group I I I B element; emission line; plasma gas The hollow anode glow discharge tube as first developed by Grimm,'V2 has been employed extensively in atomic emission spectrometry. This analytical technique has been appl led to surface analysis and to the elemental analysis of solid because cathode sputtering occurring on the sample surface results in the ejection of analyte atoms.8 Unlike the numerous reported applications little fundamental rexearch on the emission mechanisms has been performed.Several papers have pointed out that the emission lines excited hy the glow discharge depend strongly on the plasma gas 25 20 15 2 10 5 0 A12+ 20 - ..... ...._.... ,. ,...... - 3p4s 7s-6d- -7P =-7f - 4f 6s-M- -6p -\ 6f . 5s =4d= -5p -\5f t 4 N e' ~ . 5 ~ Arm 15.76 eV Fig. 1 Energy level diagram of singly ionized aluminiuni Journal of Analytical Atomic Spectrometry employed;g-" however a complete wavelength table of the glow discharge spectrum has not been published. It is therefore worthwhile classifying the emission lines for analytical use so that the analytical line and the operating conditions of the excitation source can be properly determined. It has been reported that in Grimm glow discharge particular ionic lines are selectively excited through (quasi-)resonance charge-transfer ~ollisions'~ between a ground state ion of the plasma gas (G) and a ground state atom of the sample (M) G + + Mg-+Gg+(M')*.This reaction produces a highly populated excited state of the sample (M+)* which subsequently leads to the emission of an ionic line. It should be noted that such excitations occurring in the glow discharge plasma are primarily dependent on the properties of the plasma gas such as the ionization energy which implies that selectively excited ionic lines might be found for various elements. h cn C 3 - .- 2 2 e .- c 4- .- (d v cn Q c c 0 fn fn W c. .- .- .- E n 357 358 359 360 361 Wavelengthhm Fig. 2 Spectral scans of A1 I1 emission lines in the wavelength range 357-361 nm excited by an argon (a) an argon-helium mixed gas (b) and a neon plasma (c). Discharge conditions Ar 4.0 x lo2 Pa; 500 V (a) Ar 4.0 x lo2 Pa plus He 5.3 x lo2 Pa; 500 V (b) and Ne 5.3 x lo2 Pa; 600 V (c) Journal of Analytical Atomic Spectrometry October 1996 Vol.1 1 (957-966) 957Table 1 Observed emission lines of aluminium and their assignment Wavelength/nm Assignment Reduced intensity* in vacuo I1 162.562 I1 167.079 I1 171.944 I1 172.124 I1 172.127 I1 172.495 I1 172.498 I1 175.062 I1 176.010 I1 176.197 I1 176.395 I1 176.581 I1 176.773 I1 182.857 I1 183.283 I1 183.481 I1 185.592 I1 185.802 TI 185.999 I1 186.231 I1 192.996 I1 193.237 I1 193.449 I1 193.470 I1 193.691 I1 193.925 I1 196.146 I1 196.265 I1 196.269 I1 196.273 I1 199.053 I1 207.466 I1 209.492 I1 209.540 I1 209.545 I1 209.571 IT 209.576 I1 209.579 I 214.623 1215.138 I 215.140 I 216.948 I 217.475 I 217.477 I1 219.329 I1 219.488 I1 219.493 I1 219.608 I1 219.614 I1 219.618 I 220.535 I 221.075 I 221.082 I 226.416 I 226.979 I 226.992 I1 232.490 I1 232.615 I1 232.620 I1 232.712 I 236.778 I 237.385 I 237.408 I 238.080 in air 207.400 209.425 209.473 209.478 209.504 209.509 209.5 12 214.555 21 5.070 21 5.072 2 16.880 2 17.407 2 17.409 2 19.260 2 19.4 19 219.424 2 19.539 2 19.545 219.549 220.466 22 1.006 22 1.01 3 226.346 226.909 226.922 232.419 232.543 232.549 232.641 236.706 237.3 13 237.336 238.007 Upper (eV) 5 s 'So (15.047) 3p 'P (7.4205) 3d 3D (11.847) 3d 3D (11.847) 3d 3D2 ( 11.846) 3d 3D (11.846) 3d 3D3 (11.846) 7f IF3 (17.680) 3p 3P2 (11.688) 3p 3P1 (1 1.672) 3p 3P (1 1.688) 3P 3P (1 1.665) 3p 3P (1 1.672) 3p4s jP (18.097) 3p4s 3P (18.081) 3p4s 3P0 (18.073) 4s 3S (11.316) 4s 3S (11.316) 6f 'F3 (17.264) 4s 3S (11.316) 3p4s 'P ( 18.097) 3p4s 3P1 (18.081) 3p4s 3P (18.097) 3p4s 3P (18.081) 3p4s 3P0 (18.073) 3p4s 3P (18.081) 8f 3F (17.993) 9f 3F3 (18.163) 9f 3F3 (18.163) 9f 3F2 (18.163) 3d ID (13.649) 5f 'F3 (16.574) 7f 3F (17.764) 7f 3F3 (17.763) 7f 3F3 (17.763) 7f 3F (17.762) 7f 3F2 (17.762) 7f 3F (17.762) 9d 2D3/2 (5.7767) 9d 'D,/ (5.7768) 9d 2D3/ (5.7767) 8d 'D3/ (5.7148) 8d 2Ds,2 (5.7148) 3p3d 3F (17.499) 3p3d 3F3 (17.495) 3p3d 3F3 (17.495) 3p3d 3F2 (17.492) 3p3d 3F (17.492) 3p3d 3F2 (17.492) 7d 'Ds12 (5.6220) 8d 2D3,2 (5.7148) 7d 'D3/2 (5.6218) 7d 'D3/2 (5.6218) 6d 2D3/ (5.4758) 6d 2D,I (5.4761) 6f 3F (17.179) 6f 3F3 (17.176) 6f 3F3 (17.176) 6f 3F (17.174) 5d 'D5/ (5.2367) 6s 'SlI2 (5.2215) 6d 'D312 (5.4758) 5d 'D3/ (5.2362) 5d 'D3/2 (5.2362) Lower (eV) 3p P (7.4205) 3p 3P0 (4.6360) 3p 3P (4.6436) 3p 3P (4.6436) ] 3p 3P2 (4.6590) 3p 3P2 (4.6590) ] 3p ID2 (10.598) 3p 3P (4.6436) 3p 3P (4.6360) 3p 3P (4.6590) 3p 3P (4.6436) 3p 3P2 (4.6589) 4s 3S (11.316) 4s 3S (11.316) 4s 'S1 (11.316) 3p 3P0 (4.6360) 3p 3P (4.6435) 3p 'D (10.598) 3p 3P (4.6589) 3p 3P (11.673) 3p 3P0 (11.665) 3p 3P (1 1.688) 3p 3P (1 1.672) 3p 3P (11.672) 3p 3P2 ( 11.687) 3p 3P (1 1.672) 3d 3D3 ( 11.846) 3cl 3D2 ( 11.846) 3cl 3D2 ( 11.846 j 31) 'PI (7.4205) 'D (10.598) 3D3 (11.846) 3d 3D3 ( 11.846) 3d 3D2 (1 1.846) ] 3d 3D3 (11.846) 3d 3D2 ( 11.846) 3d 3D ( 1 1.847) 3s 'so (0.0000) 1 313 2P,/ (0.0000) 31) 2P,/ (0.0000) 1 1 31) 'P3/ (0.01 39) 31) 2P3/2 (0.0139) 3p 'P3/2 (0.0139) 31) 'P312 (0.0139) 3d 3D3 (1 1.846) 3d 3D3 ( 11.846) 3d 3D2 ( 11.846) ] 3d 3D3 ( 11.846) 3d 3D ( 11.846) 3d 3D ( 11 347) 3d 2P,/ (0.0000) 3p ZP,/ (0.0000) 1 1 1 31) 'P3/2 (0.0139) 3p 'P3/ (0.0139 j 31) 'P3/ (0.0139) 3p 2P3/2 (0.0139) 3d 3D3 (11.846) 3d 3D3 (1 1.846) 3d 3D ( 11.846) 3tl 3D3 ( 11.846) 313 2P112 (0.0000) 31) 'P3/2 (0.0139) 3p 'P3/ (0.0139) 313 'P3/2 (0.01 39) Art < 1 200 20 25 45 < 1 5 5 20 10 5 < 1 < 1 < 1 6 10 < 1 15 < 1 < 1 < 1 < 1 < 1 <1 < 1 ArT[ 10 < 1 < 1 < 1 20 40 30 60 <1 < 1 ArT[ 50 100 80 130 < 1 < 1 < 1 70 y 8 Ar-Hef 30 520 170 5 10 850 15 60 50 210 60 65 160 120 50 100 3 10 45 5 10 45 (40) 130 55 65 30 20 340 130 80 40 30 60 65 140 110 90 (80) 140 250 220 370 8 8 8 220 330 55 25 Ne§ 310 1500 540 1600 2800 <1 120 80 380 70 95 < I < 1 < l 330 970 < I 1600 100 < 1 < 1 < 1 < 1 < 1 < 1 <1 1100 < 1 <1 < 1 5 8 9 15 < 1 < 1 < 1 15 30 25 45 < 1 < 1 < 1 25 40 10 5 * Normalized per unit amount of the sputtered sample.t Discharge conditions Ar 4.0 x 10' Pa; 400 V; 42.2 mA. $ Discharge conditions Ar 4.0 x 10' Pa plus He 5.3 x 10' Pa; 450 V; 38.6 mA. 0 Discharge conditions Ne 5.3 x 10' Pa; 700 V; 20.8 mA. 7 Not estimated due to overlapping with gas emission lines. 11 Standard for estimating the reduced intensities. 958 Journal of Analytical Atomic Spectrometry October 1996 Vol. 11Table 2 Observed emission lines of aluminium and their assignment Wavelenrrthl Assignment Reduced intensity* (nm) (in iir)' 1236.706 1237.314 I1 247.526 I 236.798 1257.509 I1 263.152 I1 263.769 I1 263.818 I1 263.825 I1 263.853 I1 263.860 I1 263.867 1265.248 I 266.039 I1 266.916 I1 281.618 I1 299.547 I1 299.817 I 308.215 I 309.270 I 309.284 I1 358.660 I1 358.699 I1 358.713 I1 358.725 I1 358.751 I1 364.926 I1 365.099 I1 365.500 I1 370.334 I1 373.195 I1 373.377 I1 386.618 I 394.401 I 396.153 I1 466.296 I1 559.348 I1 585.387 I1 586.176 I1 618.143 II 618.220 I1 618.335 I1 620.175 I1 622.648 I1 623.153 I1 624.320 I1 633.579 IT 681.708 I1 682.313 I1 683.712 I1 692.042 I1 704.205 Upper (ev) 5d ,D3/2 (5.2362) 5d ,D5! (5.2367) 5p 'P (15.605) 4f 'F3 (15.308) 5f 3F (16.545) 5f 3F3 (16.544) 5f 3F3 (16.544) 5f 3F (16.544) 5f 3F (16.544) Sf 3F (16.544) 5s 'S1/ (4.6728) 5s 2S112 (4.6728) 3p 3P1 (4.6436) 4s 'So (11.822) 6d 3D, (17.210) 6d 3D,,2 (17.210) 3d 'D,,' (4.0214) 4d 'D3/2 (4.8265) 4d 'D5,2 (4.8271) 3d ,D5/2 (4.0216) 3d ,D3/2 (4.0214) 4f 3F4 (15.302) 4f 3F3 (15.302) 4f 3F3 (15.302) 4f 3F (15.301) 4f 3F (15.301) 5d 3D1 (16.468) 5d 3D1,2 (16.468) 5d 3D2 (16.468) 5d 'D2 (16.603) 6s 3S1 (16.392) 6s 3S1 (16.392) 6s 'So (16.462) 4s 'Sl/ (3.1426) 4s 2Sl/z (3.1426) 'P (13.256) 3 'D (15.472) 6f 3F4 (17.179) 6f 3F3 (17.176) 6g 3G3 (17.307) 6g 3G3,4 (17.307) 6g 3G3.4,5 (17.307 6g 'G (17.307) 4d 3D (15.062) 4d 'D1.(15.062) 4d 3D1,2,3 (15.062 5p 'P (15.605) 5s 3S1 (14.889) 5s 3S1 (14.889) 5s 3S1 (14.889) 5s 'So (15.047) 4p 3P (13.076) Lower (eV) 3p (0.0000) 3p 2P1 (0.0000) 3p 'P3/2 (0.0139) 3p 'D (10.598) 3p 'P, (0.0139) 3p 'D (10.598) 3d 3D3 (11.846) 3d 'D3 (1 1.846) 3d 3D2 (1 1.846) 3d 3D3 ( 1 1.846) 36 3D (1 1.846) 3d 3D1 (1 1.847) 3p 2P3/2 (0.0139) 3p 'P (7.4205) 4p 3P1 (13.073) 4p 'P (13.076) 3P ,p1/2 ( 0 .~ 0 ) 3s 'so (0.0000) 3p 2P1 (0.0000) 3p ,P3/2 (0.0139) 3p ,P3/2 (0.0139) 3d 3D3 ( 11.846) 3d 3D3 (1 1.846) 3d 3D2 (1 1.846) 3d 3D3 (1 1.846) 3d 'D (1 1.847) 4p 3P (13.071) 4p 3P1 (13.073) 4p 3P2 ( 13.076) 4p 'P1 (13.256) 4p 3P0 (13.071) 4p 3P1 (13.073) 4p 'PI (13.256) 3p 'D (0.0000) 4p 'P1 (13.256) 4d 3D3 ( 15.062) 4d 3D2 (15.062) 4f 3F2 (15.301) 4f 3F3 (15.302) 4f 3F4 (15.302) 4f 'F3 (15.308) 4p 3P0 (13.071) 4p 3P (13.073) 4p 3P (13.076) 3d 'D (13.649) 4p 3P0 (13.071) 4p 3P1 (13.073) 4p 3P2 (13.076) 4p 'P (13.256) 4s 3S1 (11.316) 3p 2P1/2 (0.0000) 3p 'P3/2 (0.0139) Art 60 100 II <1 140 230 < 1 <1 1 1 <1 <1 <1 70 140 < 1 30 <1 < 1 2700 4100 <1 < 1 < 1 <1 <1 < 1 < 1 <1 < 1 8700 9900 < 1 <1 <1 <1 < 1 <1 <1 < I <1 <1 <1 < 1 <1 <1 < 1 <1 ArB Arll Ar-He$ 190 320 15 540 890 280 45 30 30 30 230 480 2 250 4 4 7300 14 000 1800 1300 1100 45 160 90 20 30 45 12 OOO 18 000 60 55 8 5 8 8 15 8 40 120 (230) 3 10 20 30 35 Ne§ 35 55 200 80 150 1700 <1 <1 <1 <1 40 75 40 180 <1 < 1 1800 3100 14000 10 OOO 7900 1 3 3 3 <1 <1 <1 4100 5300 2200 840 < 1 < 1 < 1 <1 <1 1100 3300 5800 140 50 160 320 700 840 Nen * Normalized per unit amount of the sputtered sample.t Discharge conditions Ar 4.0 x 10' Pa; 400 V; 47.2 mA. 1 Discharge conditions Ar 4.0 x lo2 Pa plus He 5.3 x 10 Pa; 450 V; 42.0 mA. 6 Discharge conditions Ne 5.3 x 10 Pa; 700 V; 23.6 mA. f Not estimated due to overlapping with gas emission lines. // Standard for estimating the reduced intensifier. In this paper the emission lines of the Group IIIB elements aluminium gallium and indium emitted by Grimm glow discharge plasmas were investigated. The emission lines were classified according to their transition schemes and the relative intensities were compared for several different plasma gases uiz. argon neon argon-helium and neon-helium mixed gases. A common feature of the Group IIIB elements is that the second ionization energy is relatively high whereas the first ionization from the neutral atom takes place easily.Hence there are distinct excited levels belonging to the various electron configurations leading to spectra rich in the ionic emission lines. It has been reported that the hollow cathode laser based on Al I1 emission lines can be pumped by energy transfer from neon ions. Steers and Leis15 have indicated that the intense A1 I1 lines assigned to the 3s4f-3s3d transition can be excited principally through charge-transfer collisions in which ground state ions of neon are involved. Previous studies have also shown that for the Group IIIB elements the spectra obtained with the glow discharge might be different from those obtained with other sources and might depend on the plasma gas used. Hence complete wavelength tables are required not only for analytical applications but also for investigations into the excitation mechanisms.EXPERIMENTAL The structure of the glow discharge tube employed here has been described el~ewhere.'~ It was constructed according to the original model reported by Grimm.' The inner diameter Journal of Analytical Atomic Spectrometry October 1996 Vol. 11 959a 1 182 183 184 185 186 187 Wavelengthhm Fig. 3 Spectral scans of A1 I1 emission lines in the wavelength range 182-187 nm excited by an argon (a) an argon-helium mixed gas (b) and a neon plasma (c). Discharge conditions Ar 4.0 x 10’ Pa; 550 V (a) Ar 4.0 x lo2 Pa plus He 5.3 x 10’ Pa; 550 V (b) and Ne 5.3 x 10’ Pa; 700 v (c) A 0- 200 400 600 80 2 0 0 4 0 0 6 0 0 8 0 0 of the hollow anode was 8.0mm and the distance between the anode and cathode was adjusted to be 0.2-0.3mm.The discharge tube was evacuated down to about 1.3 Pa (1 x lo- Torr) and then the plasma gas was introduced. High- purity argon (99.9995%) neon (99.99%) and helium (99.9999%) were used. Flow control of the plasma gas was carried out with a ball (on/off) valve and a needle valve which were inserted in each gas line. The partial pressure of each gas was regulated with the needle valve and read on a Pirani gauge at the gas inlet of the tube when the ball valves of the other gas lines were closed. The scales of the gauge had been calibrated for each pure gas. The discharge power was supplied with a dc power supply device (Model PAD lK-0.2L Kikusui Electronics Tokyo Japan).All measurements were conducted in constant-voltage mode. A Fastie-Ebert mounting spectrograph (Model GE-340S Shimadzu Kyoto Japan) equipped with a photomultiplier tube (Model R-955 Hamamatsu Photonics Tokyo Japan) was employed to measure the spectra in the wavelength region 230-800 nm. The focal length is 3.4 m. The grating has 1200 grooves mm-’ and a blaze wavelength of 300 nm. The emission spectra in the 160-245 nm wavelength range were recorded on an Eagle-mounting vacuum spectrometer (Model EGV-200 Shimadzu) equipped with a CaF window and a photomultipl- ier tube (Model R-l66UH Hamamatsu Photonics). The focal length is 2.0 m. The concave grating has 1200 grooves mm-l and a blaze wavelength of 170nm. The spectrometer was evacuated to about 1.3 Pa with a rotary vacuum pump.The intensity data were recorded at scan rates of 0.008 nm s-l (GE-340s) and 0.010 nm s-l (EGV-200) with an analogue chart recorder. (b) I (8) Al I1 624.32 nm I ! ’ 0 L o -0 - -0 _I L.-8.-:-;; 200400600800 2 0 0 4 0 0 6 0 0 8 0 Discharge voltageN Fig. 4 Discharge voltage dependence of the discharge current (a) the sputtering rate (b) and of the emission intensities of A1 I1 281.618 nm (c) A1 I1 358.660 nm ( d ) A1 I1 624.320 nm (e) and A1 I 396.153 nm (f). Plasma gases Ar 4.0 x lo2 Pa (circles) and Ne 5.3 x 10’ Pa (squares) 960 Journal of Analytical Atomic Spectrometry October 1996 Vol. 1115 t 3 - 2 ' 1 ' 0 A (e) Al II 209.49 nm A / A/ ,/ A' A A - o-mo--@ -I . 0 400 600 800 1000 i .1" i i ,m' 11' ,i i I I d 60 40 20 0 d . 6 8 - 400 600 800 1000 100 t 0 v) .- .% 20 i c Q Z O 400 600 8CiO loo0 Discharge voltageN Fig.5 Discharge voltage dependence of the discharge current (id) and of the emission intensities of A1 I1 162.562 nm (b) A1 I1 167.079 nm (c) A1 I1 172.492 nm (d) and A1 I1 209.492 nm (e). Plasma gases Ar 4.0 x 102 Pa (circles) Ar 4.0 x lo2 Pa plus He 5.3 x ld Pa (triangles) and Ne 5.3 x lo2 Pa (squares) Table 3 Observed emission lines of gallium and their assignmeni Wavelength/nm Asbignment Reduced intensity* in vacuo I1 153.551 I1 153.631 I1 166.937 I1 169.581 I1 179.944 I1 181.399 I1 184.526 I1 209.135 1219.603 1223.661 I 225.989 1229.487 1229.855 1233.894 1233.923 1237.201 I 241.942 I 245.082 I 250.094 in air 209.069 219.534 223.592 225.989 229.416 229.784 233.822 233.851 237.129 241.869 245.008 250.019 Upper (eV) 4d 3D3 (14.118) 4d 'D2 (14.1 14) 4d ID2 (13.355) 4d 'D (13.355) 5s (12.763) 5s ?St (12.763) 5s 3S1 (12.763) 4p 3P1 (5.9283) 9s 2s1/2 (5.6457) 9s 2s1/2 (5.6457) 7d 2D5/2 (5.5886) 6d 'D312 (5.4025) 6d 'D5/2 (5.4032) 6d 'D3p (5.4025) 8s 2S,12 (5.4963) 7s 2Sl/2 (5.2269) 7s 2S112 (5.2269) 5d 'DU2 (5.0598) 5d 'D3/2 (5.0588) Lower (eV) 4p 3P2 (6.0441) 4p 3P2 (6.0441) 4p 3P (5.9283) 4p 3P2 (6.0441) 4p 'Po (5.8730) 4p 3P1 (5.9283) 4p 3P2 (6.0441) 4s 'So (O.oo00) 4P 2p1/2 (O.oo00) 4P 2pl (O.oo00) 4P %/2 (O*ooOo) 4P (O.oo00) 4p 'P3/2 (0.1024) 4p 'P3/2 (0.1024) 4p 2P3/2 (0.1024) 4p 2P3/2 (0.1024) 4p 'P3/2 (0.1024) 4p 'P3/2 (0.1024) 4p 'P312 (0.1024) Art <1 < 1 < 1 < 1 < 1 <1 <1 780 10 10 10 30 15 45 6 10 15 80 loon Ar-He$ 15 < 1 < 1 30 100 400 640 1500 70 50 30 120 40 190 25 30 65 360 460 N4 160 40 30 220 1400 4200 6800 6OOo <1 < 1 < 1 5 5 20 1 5 10 45 50 * Normalized per unit amount of the sputtered sample.t Discharge conditions Ar 4.0 x 102 Pa; 450 V; 28.0 mA. $ Discharge conditions Ar 4.0 x I d Pa plus He 5.3 x lo2 Pa; 501 V; 26.8 mA. Q Discharge conditions Ne 5.3 x lo2 Pa; 800 V; 12.3 mA. Standard for estimating the reduced intensities. Aluminium plates (99.99% purity) tin-gallium alloy blocks (containing about 20% m/m Ga) and tin-indium alloy blocks (containing about 20% m/m In) were prepared as the cathode samples. It is particularly difficult to handle gallium metal as the sample because of its low melting-point. However pro- longed stable discharges could be attained for the alloy samples.The spectral analysis of gallium or indium emission lines was not significantly affected by the appearance of emission lines Journal of Analytical Atomic Spectrometry October 1996 VoL 11 961100 \ ‘. \ \ \ a\ - ‘h 7’ \A/o ’b Helium partial pressure/l O2 Pa - I - - . - (h) A1 I1 209.49 nrn 7 - 4 - / / m 2 - L/ 17’’ // // . / / A * . . ‘ - Fig. 6 Variations in the discharge current (a) the sputtering rate (b) and in the emission intensities of A1 I1 162.562 nm (c) A1 I1 167.079 nm ( d ) Al I1 172.492 nm (e) A1 I1 186.231 nm (f) A1 I1 207.466 nm (8) and A1 I1 209.492 nm (h) as a function of helium partial pressures added to an argon (circles) or a neon (squares) plasma. Discharge conditions Ar 4.0 x 10’ Pa; 700 V and Ne 5.3 x 10’ Pa; 800 V of tin because only a few such lines were observed in the spectrum. of the sputtered sample (the sputtering rate) so that they can be compared for different plasma gases.From the viewpoint of the optical transition the observed A1 I1 lines can be RESULTS AND DISCUSSION Aluminium Table 1 summarizes the A1 I and A1 I1 emission lines measured with the vacuum spectrometer and Table 2 the aluminium lines measured with the air-path spectrometer. The first and second columns give the wavelength values which were calcu- lated from the energy level data in ref. 20 and their assignments respectively. The remaining columns provide the emission intensities observed with argon gas an argon-helium mixture and neon gas. The intensities are normalized per unit amount classified -into the following groups 3s3p + 3s2 (resonance transition) 3sns (n =4 5) -+ 3s3p 3P2 + 3s3p 3snd (n = 3 4) 4 3s3p 3snf (n =4 5 6 .. .) + 3 ~ 3 d 3p4s 4 3s4s or 3p2 3p3d + 3s3d Fig. 1 shows a schematic energy diagram of singly ionized aluminium together with the ionization potentials of argon neon and helium. 962 Journal of Analytical Atomic Spectrometry October 1996 Vol. 1 1I 25 - 20 2 15 10 5 Ar Ne He In2+ - 61 -51 r 7s -6d- -7P - 41 - as -7d- - 5p2=5d= - ==6P 6s '-" Ar + T - IErn 21 .! 5P - 15 76 eV I In+ .. . . 7 5.78 ev 0 - i ' A- Fig. 8 Energy level diagram of singly ionized indium Fig. 7 Energy level diagram of singly ionized gallium Table 4 Observed emission lines of gallium and their assignme!it Wavelength/ (nm) (in air) I 241.868 I 245.008 I 250.019 I 250.072 I1 251.351 I1 270.052 I 271.966 I1 278.036 I 287.424 I 294.364 I 294.417 I1 296.940 I1 297.081 I1 297.160 I1 297.461 I1 337.467 I1 337.581 I1 347.039 I1 370.581 I1 374.488 I1 392.462 1403.297 I 417.203 I1 425.104 I1 425.376 I1 425.557 I1 426.174 I1 633.418 I1 641.960 I1 645.650 Assignment Reduced intensity* Upper (eV) 7s 2Sl (5.2269) 5d 2Dsiz (5.0598) 5f 'F3 (18.286) 4d 'D2 (13.355) 6s 'Sli2 (4.6598) 5s 'So (13.233) 4d 2DS,2 (4.3131) 5f 3F2 (18.285) 5f 'F3 (18.286) 5f 3F2 ( 18.285) 5f 3F2,3 (18.285) 4f 'F3 ( 17.028) 4f 3F2 (17.027) 6d 3D3 (18.291) 7s 3S1 (18.038) 7s 3S1 (18.038) 7s 'So (18.103) 5s 2S1/2 (3.0733) 5s 2 S l i 2 (3.0733) 4f 3F2 (17.027) 4f 'F3 ( 17.028) 4f 3F2.3 (17.027) 4f 3F2,3,4 (17.027) 5d 2D3i2 (5.0588) 5d 'D3/2 (5.0588) 4d 'D3,2 (4.3 123) 4d 2D3/2 (4.3123) 5p 3P2 (14.720) 5p 3P1 (14.694) 5p 'Po (14.683) ~ Lo..ver (eV) 4p 2P /2 (0 1024) 4p 'P (0.1024) 4p 2P ,2 (0.1024) 4d 'D I (13.355) 4p 'P (8.7654) 4p 'P ,2 (0.1024) 4p 'P (8.7655) 4p 'P (0.1024) 4p 'P ,2 (0.1024) 4d 3D (14.1 11) 4d 3D (14.114) 4d 3D (14.1 14) 4d 3D (14.118) 4d 'D (13.355) 4d 'D (13.355) 5p 3P (14.720) 5p 3P (14.694) 5p 3P (14.720) 5p 'P (14.945) 4p 'P (0.1024) 4d 3D (14.111) 4d 3D (14.114) 4d 3D (14.114) 4d 3D (14.1 18) 5s 3S1 (12.763) 5s 3S1 (12.763) 5s 3S1 (12.763) 4p 2P ,' (O.OOO0) 4p 2P 2 (0.0000) 4p 'P 2 (0.0000) Art 3 45 15 < 1 < I 110 < 1 1500 2300 480 <1 < I < l < 1 < 1 < 1 < 1 < 1 < 1 < I 6400 9500 < 1 < 1 <1 < 1 <1 < 1 < 1 loon Ar-He 20 220 480 65 30 35 420 40 4300 7500 1400 95 25 120 220 200 55 55 30 60 200 12 000 18 OOO 500 150 500 lo00 80 20 3 NeS 3 30 70 7 2 400 8 1800 900 1600 280 < 1 < 1 < 1 < 1 < 1 < 1 < 1 < 1 < 1 < 1 2800 4800 < 1 < I < 1 < 1 3200 300 60 * Normalized per unit amount of the sputtered sample.t Discharge conditions Ar 4.0 x lo2 Pa; 450 V; 30.2 mA. $ Discharge conditions Ar 4.0 x lo2 Pa plus He 5.3 x 10' Pa; .;OO V; 29.2 mA. 9 Discharge conditions Ne 5.3 x lo2 Pa; 800 V; 13.8 mA. Standard for estimating the reduced intensities. Fig. 2 indicates spectral scans in the wavelength range 357-361 nm measured with argon gas alone (a) with an argon- helium mixed gas (b) and with neon gas (c). In the argon- excited plasma no A1 11 lines appear in the spectrum consisting of several emission lines of argon ions.However in addition to neon lines at least three A1 I1 lines uiz. 358.660 358.713 and 358.751 nm lines are observed when using neon gas. The A1 I1 lines result from the 3s4f-3s3d transition (see Tables 1 and 2). The addition of helium to the argon plasma enables these A1 I1 lines to be emitted although the intensities are weaker than those in the neon plasma. Fig. 3 shows spectral scans in the wavelength range 182-187 nm. Intense triplet A1 I1 lines which are attributed to transitions from the 3s4s excited level (1 1.316 eV) appear in the neon-excited spectrum. Journal of Analytical Atomic Spectrometry October 1996 Vol. 1 1 963Table 5 Observed emission lines of indium and their assignment Wavelength/nm Assignment Reduced intensity* in vucuo in air I1 157.153 I1 158.645 I1 165.743 I1 167.199 I1 167.412 I1 170.007 I1 170.256 I1 171.662 I1 177.066 I1 177.484 I1 184.248 I1 193.063 I1 193.633 I1 196.672 I1 197.746 11 207.935 207.869 I1 230.686 230.6 15 Upper (eV) 6d 'D (15.707) 5p 'P (7.8150) 7s ' S o (15.295) 5d 'D (12.656) 5p2 3P2 (13.087) 5d 3D2 (12.667) 5d 'D (12.656) 5p2 3P0 (12.567) 5d 3D3 ( 12.684) 5d 3D2 (12.667) 5p2 'D (12.103) 5p2 'D (12.103) 6s 3S1 (11.644) 5d 'D ( 14.1 19) 6s 3S1 ( 1 1.644) 6s 'S (11.644) 5p 3P (5.3745) Lower (eV) 5p 'P (7.8150) 5p 'P (7.8150) 5p 'Po (5.2413) 5p 3P (5.6817) 5p 'P (5.3745) 5p 3P1 (5.3745) 5p 'P (5.3745) 5p 3P (5.68 17) 5p 3P (5.6817) 5p 'P (5.3745) 5p 'P (5.6817) 5p 3P0 (5.2413) 5p 'P (7.8150) 5p 'P (5.3745) 5p 'P (5.6817) 5s ' S o (O.oo00) 5s ' S o (O.oo00) Art < 1 4 < 1 < 1 < 1 < 1 < 1 < 1 2 < 1 < 1 < 1 < 1 < 1 1 2 loql Ar-He$ 1 7 < 1 3 1 7 2 < I 20 3 < I < 1 1 5 4 8 130 Ne$ 8 50 25 15 3 40 10 5 110 15 4 15 6 20 15 35 190 * Normalized per unit amount of the sputtered sample.t Discharge conditions Ar 4.0 x 10 Pa; 450 V; 31.6 mA. $ Discharge conditions Ar 4.0 x 10 Pa plus He 5.3 x 10' Pa; 500 V; 26.8 mA. 9 Discharge conditions Ne 5.3 x 10 Pa; 800 V; 11.0 mA. 1 Standard for estimating the reduced intensities. Such triplet A1 I1 lines are also observed with argon gas although the intensities are much lowered compared with those in the neon plasma. It is further noted that in the argon- helium spectrum there is another group of A1 I1 emission lines uiz.182.857 183.283 and 183.481 nm lines which are not emitted by the neon or argon plasma. These A1 I1 lines result from the 3p4s-3~4~ transition and the upper levels have excitation energies of more than 18 eV (see Tables 1 and 2). It is possible to observe these A1 I1 lines also in neon-helium mixed gas plasmas which implies that helium is principally responsible for the excitations. Fig. 4 shows the voltage dependence of the discharge current (a) the sputtering rate (b) and the emission intensities' of some aluminium lines (c)-(j') when argon (circles) and neon (squares) are employed as the plasma gases. It was found that both the sputtering rate and the discharge current were raised with increasing discharge voltage and that at the same voltage both were greater in the argon-excited plasma than in the neon-excited plasma.As indicated in Fig. 4 ( n the intensity of the A1 I 396.153 nm line emitted by the argon plasma is always larger than that emitted by the neon plasma which roughly corresponds to changes in the sputtering rate. However the observed intensity variations of the A1 I1 lines are unlikely to be explained by differences in the discharge conditions between argon and neon gas. The intensities of the A1 I1 358.660nm [Fig. 4(d)] and A1 I1 624.320 nm [Fig. 4(e)] lines in the argon plasma are almost independent of the applied voltage whereas those in the neon plasma increase with increasing voltage. The intensity of the A1 I1 281.618 nm [Fig.4(c)] line exhibits a similar voltage-dependence for both plasmas but the intensity emitted by the argon plasma is weaker.Fig. 5 shows intensity variations of some A1 I1 lines measured in the vacuum ultraviolet wavelength region. The results obtained with an argon-helium mixed gas are also plotted. The voltage-current characteristics are different from those in Fig. 4 although the same discharge conditions were employed; this is principally because the structure of the glow lamp is altered in order for it to be set in the vacuum spectrometer. Regardless of the discharge voltages the A1 I1 209.492 nm line [Fig. 5(e)J which results from the 3s7f-3s3d transition is not significantly excited either in the argon or the neon plasma indicating that this line is observable only with the argon- helium mixed gas.On the other hand the intensity of the A1 I1 964 Journal of Analytical Atomic Spectrometry October 1996 167.079 nm line [Fig. 5(c)] which is attributed to a resonance transition exhibits a similar voltage dependence for all the plasma gases employed even though the intensity in the neon plasma is generally greater than in the other plasmas. The intensity variation of the A1 I1 162.562 nm line [Fig. 5(b)] is similar to that of the A1 I1 358.660nm or A1 I1 624.320nm line [Fig. 4(d) and (e)]. Fig. 6 shows variations of the discharge current (a) the sputtering rate (b) and the emission intensities of the vacuum A1 I1 lines (c)-(h) as a function of helium partial pressures added in an argon (circles) and a neon (squares) plasma. It was found that both the sputtering rate and the discharge current decrease with increasing amounts of helium in the plasmas.Such a decline in the discharge conditions might lead to a decrease in the emission intensities. Nevertheless the addition of helium leads to complicated changes in the A1 I1 intensities depending on the nature of the lines. The intensities of the A1 I1 207.466nm line [Fig. 6(g)] and the A1 I1 209.492 nm line [Fig. 6(h)] correspond to the amount of helium added in a straightforward manner which is hardly dependent on the type of matrix gas (argon or neon). Such results imply that helium species in the plasmas mainly contribute to ioniz- ation and excitation processes for the corresponding excited states of aluminium ions. The A1 I1 182.857 nm line A1 I1 193.449 nm line etc.exhibit similar behaviour regarding the addition of helium. On the other hand the addition of helium to the neon plasma causes a significant decrease in the intensity of the A1 I1 162.562 nm line [Fig. 6(c)] whereas the intensity gradually increases in the argon plasma. Similarly the intensity of the A1 I1 172.495 nm line [Fig. 6(e)] or the A1 I1 186.231 nm line [Fig. 6 0 1 decreases with the neon-helium plasmas. When the neon-helium plasma includes helium partial pressures of 5.6 x lo2 Pa the intensity of the A1 I1 162.562 nm line is reduced by a factor of 8 and that of the A1 I1 186.231 nm line by a factor of 2 thus differing for each A1 I1 line analysed. The addition of helium to argon plasmas generally enhances the A1 I1 intensities regardless of the reduction of the sputtering rate and many more A1 I1 lines are observable compared with excitation with argon gas alone.It can be concluded from these results that additional excitations caused by helium species principally determine the spectral pattern of A1 I1 emission lines. However in a neon plasma the addition of helium results in intensity decreases of the A1 I1 lines which Vol. 11Table 6 Observed emission lines of indium and their assignment Wavelength/ nm (in air) I1 230.615 1252.137 I 252.299 I 256.015 I 260.175 I1 268.312 1271.027 1271.394 I1 274.974 1275.388 1277.535 1283.691 1285.812 I1 289.018 1293.263 I1 294,104 I 295.700 I 303.936 I1 313.861 I1 314.671 I1 315.835 I 325.609 I 325.856 I1 333.848 I1 379.528 I1 383.468 I1 384.220 I1 384.293 I1 390.194 1410.176 1451.130 I1 463.821 I1 464.468 I1 465.571 I1 465.680 I1 467.355 I1 468.100 I1 468.473 I1 585.318 I1 590.329 I1 591.517 I1 591.867 I1 609.624 11 611.601 Assignment Reduced intensity* Upper (eV) 5p ,P1 (5.3745) 7d 2D5/2 (5.1901) 7d 2D3/2 (5.1869) 6d 2D3,z (4.8413) 8s 2S1/2 (5.0382) 6f ,F4 (17.303) 6d ,D5/2 (4.8475) 6d 2D3/2 (4.8413) 5f 'F (16.611) 5p 4P,/ (4.4659) 5p 4P,i (4.6433) 5p 4P,/2 (4.3366) 5p ID2 (12.103) 6s ' S o (12.029) 5p 4P,/ (4.4659) 5f 'F2 (16.606) 5f ,F3 (16.606) 5f ,F4 (16.608) 7s 2s1 (4.5007) 7s 2s1/ (4.5007) 5d 2D3/2 (4.0780) 5d ,D5/2 (4.0809) 5d 'D3/2 (4.0780) 7p 'P (15.816) 7d ,D (16.709) 4f 'F (15.336) 4f ,F (15.329) 4f ,F2 (15.329) 7d 'D2 (16.787) 6s 'S1/ (3.0218) 6s 'Slj2 (3.0218) 4f ,F2 (15.329) 4f IF3 (15.336) 4f ,F3 (15.329) 4f ,F2 (15.329) 4f 'F (15.336) 4f ,F4 (15.331) 4f 'F3 (15.329) 6d ,Dl (15.465) 6d ,D (15.469) 6d ,Dl (15.465) 6d 'D (15.704) 6d ,D ( 15.476) 6d ,D (15.469) ~- Lower (eV) 5s ' S o (0.oO00) 5P ,PI/ (O.oo00) 5p ,P3/2 (0.2743) 5p 2P3/2 (0.2743) 5p ,P3/2 (0.2743) 5p 2P3/2 (0.2743) 5p ,P3/2 (0.2743) 5d ,D (12.684) 5p2 'D (12.104) 5p 2P1/ (0.0000) 5P ,p1/2 ( 0 .~ 0 ) 5P 2P1/2 ( 0 . ~ 0 ) 5p ,P3/2 (0.2743) 5p 'P (7.8150) 5p 'PI (7.8150) 5d 3D1 (12.656) 5d 3D (12.667) 5d ,D ( 12.684) 5p 'P3/2 (0.2743) 5p 2P3/2 (0.2743) 5P 2p1/2 ( 0 . 0 ~ ) 5p ,P3/2 (0.2743) 5p ,P3/2 (0.2743) 5p2 ID (12.103) 6p ,P2 (13.443) 5p2 'D (12.103) 5p2 'D (12.103) 5p2 ID (12.103) 6p 'P1 (13.610) 5d ,Dl (12.656) 5d ,D2 (12.667) 5d ,D2 (12.667) 5d 3D (12.667) 5d 3D3 (12.684) 5d ,D3 (12.684) 5d ,D (12.684) 6p 3P0 (13.348) 6p ,P1 (13.370) 6p ,P1 (13.370) 6p 'P (13.610) 6p ,P (13.443) 6p ,P2 (13.443) 5P ,p1/* ( O .O ( W 5p 'P3/2 (0.2743) Art 1w <1 <1 8 1 <1 25 3 <1 10 1 6 2 <1 25 2 3 190 <1 < 1 <1 330 160 <1 < 1 <1 <1 <1 <1 1200 1600 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 < 1 < I <1 Ar-He$. 150 8 1 50 7 1 170 25 5 45 4 30 8 10 110 20 4 780 2 2 5 2100 590 <1 3 25 <1 <3 1 1900 2800 50 6 45 4 <1 120 3 <1 1 <1 2 10 < I Ne§ 220 <1 <1 10 2 <1 20 3 1 10 1 8 4 130 15 85 4 130 <1 <1 1 280 90 15 <1 670 5 60 <1 280 500 330 50 280 30 4 740 15 10 20 3 35 Ne II 8 * Normalized per unit amount of the sputtered sample. t Discharge conditions Ar 4.0 x lo2 Pa; 450 V; 33.4 mA.$ Discharge conditions Ar 4.0 x 10 Pa plus He 5.3 x 10 Pa; 500 V; 28.4 mA. Q Discharge conditions Ne 5.3 x 10 Pa; 800 V; 11.0 mA. 7 Standard for estimating the reduced intensities. 11 Not estimated due to overlapping with gas emission lines. can be excited with the neon gas although some A1 I1 lines are first excited in the presence of helium. This is due to the reduction in the sputtering rate as indicated in Fig. 6(4. In addition the addition of helium may have an influence on the relative populations of neon species through neon-helium collisions and therefore on excitation by neon itself. As can be seen from Fig. 1 singly ionized aluminium h.as several excited states favourable for a charge-transfer collision between a ground state ion of neon and a ground state atom of aluminium.These are the 3s5s the 3s4d the 3s5p and the 3s4f excited states. It should be noted that the sensitive A1 I1 lines emitted by the neon plasma results from transitions from these excited states. The small difference in the total energies may contribute to (quasi-) resonance energy transfer thus resulting in the increased population of the corresponding excited levels. Also this reaction follows Wigner's spin rule." A characteristic of helium-containing plasmas is that A1 I1 lines with relatively high excitation energies can be observed whereas they cannot be emitted by neon and argon plasmas. The corresponding upper energy levels are deduced from the 3p4s the 3p3d or the 3s7f electron configuration. It is expected from Fig.1 that charge-transfer collisions in which helium ions are involved will be beneficial in yielding these states of the aluminium ion. It was also found in the neon-excited plasma that the A1 I1 lines originating from the states having lower energies such as the 3s34 state generally give larger emission intensities. The A1 I1 172.495 nm line is a typical example. Possible channels for their excitations are stepwise de-excitations from the excited states having higher energies. The reactions are expressed as 3s4f- 3s3d4 3s3p 3s5p+3s4s43s3p 3s5s (3s4d)+3s4p+3s4s In the neon plasma energy transfer from neon ions results in the increased population of the 3s4f state etc. The subsequent stepwise de-excitations may induce emissions of the corre- sponding A1 I1 lines. Journal of Analytical Atomic Spectrometry October 1996 Vol.1 1 965Gallium Table 3 summarizes the wavelengths of Ga I and Ga I1 emis- sion lines measured with the vacuum spectrometer and Table 4 the gallium lines measured with the air-path spectrometer. The format of Tables 3 and 4 is the same as that of Tables 1 and 2. Fig. 7 shows a schematic energy diagram of singly ionized gaIIium together with the ionization energies of argon neon and helium. Sensitive Ga I1 lines emitted by the neon plasma result from a 4s5p-4~5~ or 4s4d-4~4~ transition. The Ga I1 633.418nm line (5p 3P,) is a typical line belonging to the transition schemes. As with the excitation mechanism of the A1 I1 lines the excitation to the 4s5p and 4s4d states may be explained mainly by charge-transfer collisions between ground state ions of neon and gallium atoms.The energy surplus for instance 0.84eV in the 5p3P2 level is larger compared with those in the A1 I1 lines. This is theoretically too large for charge transfer to occur effectively; however charge transfer reactions might still be considered to explain the intensity enhancement in the neon plasma. Furthermore there are certain types of Ga I1 lines such as the Ga I1 426.174 nm line which are emitted only by the argon-helium plasma. These Ga I1 lines which are attributed to the 4s4f (or 4s5f)-4s4d transition are also observed in a neon-helium plasma. It is possible that the 4s4f and 4s5f excited states are populated through charge-transfer collisions in which ground state ions of helium are involved and sub- sequent step-wise de-excitations.CONCLUSIONS In Grimm glow discharge spectrometry the relative intensities of the emission lines of the Group IIIB elements presented in this work demonstrate that their excitations are principally determined by the nature of the plasma gas. It was found that some ionic lines are selectively emitted by a neon plasma and that some ionic lines appear only when a plasma gas containing helium is employed. These phenomena can be explained by the assumption that charge-transfer collisions between analyte atoms and plasma gas ions is the major mechanism for determination of the spectrum. Indium Table 5 summarizes the wavelengths of In I and In I1 emission lines measured with the vacuum spectrometer and Table 6 the indium lines measured with the air-path spectrometer. The format of Tables 5 and 6 is the same as that of Tables 1 and 2. Fig. 8 shows a schematic energy diagram of singly ionized indium together with the ionization energies of argon neon and helium. Sensitive In I1 lines appear in the neon-excited spectrum. These In I1 lines for example the In I1 468.100 nm or In I1 591.867 nm line are attributed to the 4s4f-5s5d or the 5s6d-5~6~ transition. As can be seen from Fig. 8 selective excitations to the 5s4f and the 5s6d states can take place through charge-transfer collisions between ground state ions of neon and indium atoms. REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Grimm W. Naturwissenschaften 1967 54 586. Grimm W. Spectrochim. Acta Part B 1968 23 443. Belle C. L. and Johnson J. D. Appl. Spectrosc. 1973 27 118. Takadoum J. Pivin J. C. Pons-Corbeau J. Berneron R. and Charbonnier J. C. Surf. Interface Anal. 1984 6 175. Teo W. B. and Hirokawa K. Surf. Interface Anal. 1988 11,421. Bengtson A. Eklund A. Lundholm M. and Saric A. J. Anal. At. Spectrom. 1990 5 563. Payling R. Jones D. G. and Gower S. A. Surf. Interface Anal. 1993 20 959. Boumans P. W. J. M. Anal. Chem. 1972 44 1219. Johansson S. Phys. Scr. 1978 18 217. Wagatsuma K. and Hirokawa K. Anal. Chem. 1985 57 2901. Steers E. B. M. and Fielding R. J. J. Anal. At. Spectrom. 1987 2 239. Wagatsuma K. and Hirokawa K. Anal. Chem. 1988,60 702. Wagatsuma K. and Hirokawa K. J. Anal. At. Spectrom. 1989 4 525. Wagatsuma K. and Hirokawa K. Spectrochim. Acta Part B 1991 46 269. Steers E. B. M. and Leis F. Spectrochim. Acta Part B 1991 46 527. Steers E. B. M. and Thorne A. P. J. Anal. At. Spectrom. 1993 8 309. Duffendack 0. S. and Black J. G. Phys. Rev. 1929 34 35. Rozsa K. Janossy M. Csillag L. and Bergou J. Phys. Lett. A 1977 63 23 1. Wagatsuma K. and Hirokawa K. Surf. Interface Anal. 1984 6 167. Moore C. E. Atomic Energy Levels NBS Stand. Circular 467 US Government Printing Office Washington DC 1958 vols. 1-111. von Engel A. Ionized Gases Clarendon Press Oxford 1965. Paper 6J039.588 Received June 5 1996 966 Journal of Analytical Atomic Spectrometry October 1996 Vol. 11