Atomic Emission Spectroscopy with Spark- or Arc · PDF fileAtomic Emission Spectroscopy with Spark or Arc Excitation 4 2 Excitation of Emissions Spectra Emission lines can not only

Atomic Emission Spectroscopy with Spark or Arc Excitation 1

Fauftritt

Atomic Emission Spectroscopy with

Spark- or Arc Excitation

Experiments with the DADOS Spectrograph and Simple Makeshift Tools

Richard Walker

Version 2.0 August 2012


Table of Contents

1 Introduction .................................................................................................................. 3

2 Excitation of Emissions Spectra ................................................................................. 4

3 Tests with Spark Excitation ........................................................................................ 5

4 Tests with Arc Excitation .......................................................................................... 24

5 The Link between the Stellar- and Laboratory Spectra ......................................... 28

6 Appendix ..................................................................................................................... 31

7 Literature and Internet .............................................................................................. 33


1 Introduction It may be little known that even amateur astronomers can generate "laboratory spectra" with very simple means. This document is intended to demonstrate simple ways, how one can detect chemical elements in certain materials, liquids and gases, with a spectrograph and easily available electronic and mechanical components.

The strongly simplified procedures, presented in this document, of course do not allow any kind of real "Analytical Chemistry". However at least qualitatively, for many substances, the main components of elements or molecules can be shown. For this purpose, the basic prin-ciple of Optical Emission Spectroscopy (OES) is applied, which is nowadays in use in com-mercial, chemical analysis equipments. For this, a sample is excited to such an extent, that it emits a spectroscopically analysable plasma. The wavelength range of professional appli-cations reaches mostly from about 2,000 – 10,000 Å. With simple amateur means, analo-gous to the Astro Spectroscopy, some 3,800 – 8,000 Å can be covered.

My personal motivation was to reproduce the link between the stellar and laboratory spec-tra - quasi on the "traces" of Fraunhofer, Bunsen and Kirchhoff. If amateurs should be en-couraged to conduct their own experiments, the purpose of this work would be fully met!

Many thanks to Urs Flükiger, who supported me with metallurgical know-how and arc weld-ing experiments, carried out with equipment and personnel of his company [62].

Richard Walker, CH 8911 Rifferswil © [email protected]


2 Excitation of Emissions Spectra Emission lines can not only be generated in stellar atmospheres and gas discharge lamps, but also under the influence of flames, electrical sparks and arcs. In the following these three principles are presented, which are very easy to implement with improvised means. In the professional area numerous other methods exist. Each of these processes is specifically suited for certain types of samples.

2.1 Flame Excitation

This is certainly the oldest and also sim-plest method. Generally known is the “legendary” Bunsen Burner, which is omnipresent in most of the laboratories (picture from [20]). Further also the fact that common table salt, strewed into the flame, generates the characteristic yel-low emission of the sodium doublet D1, D2. Also many gases can be analised this way. The intense combustion of hy-drocarbon gases, such as Butane, pro-duces directly the emission of the fa-mous molecular Swan bands (C2), as shown in [13].

The disadvantages of this method are the relatively low excitation energy, the latent fire hazard, and the damage or even destruction of the sample.

2.2 Spark Excitation

This is also a very old method, already applied by Bunsen and Kirchhoff [20]. An electric spark generates plasma from a tiny fraction of "vaporized" electrode material, as well as from the struked gas mixture between the electrodes. Thereby temperatures of some 4000 – 5000 K are achieved, corresponding to the stellar atmospheres of the spectral class K. This excitation energy is sufficient even for non-destructive analysis of metals and alloys. Performed with improvised means like piezoelectric push-button gas igniter eg for BBQ grill or spark induc-tors, this is a relatively harmless procedure.

In current commercial analytical equipment the spark excitation of the sample material usually takes place in an inert argon atmosphere. Picture: portable ana-lyse equipment with pistol grip from Angstrom Inc. The electrodes consist partly of a special graphite to minimise unwanted intrinsic lines in the spectrum.

2.3 Arc Excitation

A little more sensitive, but also more complex, is the excitation of spectra with arcs. In addi-tion, here an impairment or "consumption" of the sample material is an issue. Here mostly carbon or tungsten electrodes are used and the temperatures reach some 3000 – 4000 K, as found in the stellar atmospheres of late-K and all M-Class.


3 Tests with Spark Excitation

3.1 Equipment

The easiest and cheapest means for first attempts are certainly the already mentioned piezoelectric push-button gas igniters. I use for this purpose the „SUPI V01“, from Conrad Electronics, (spare part for BBQ gas grill). It generates per push of a button a voltage surge of about 15 kV, which allows a spark length of >1 cm.

Here mounted into a plastic mug, it is important to keep the two wires at a sufficient distance to avoid an unwanted arcing (short cut). To attenuate the noise level, the upper part of the cup is finally stuffed with foam.

The spark gap is mounted on the bottom of an upside down positioned plastic box, de-signed for electrical installations. The whole is interconnected by plugs to stay as versatile as possible. The top electrode consists of a short piece of graphite pencil lead, which is at-tached with a lustre terminal to a thick, stiff but still formable copper wire. Such a graphite electrode is of course only an imperfect substitute for so-called "spectrally pure" graphite. Anyway relatively "soft" molecular C2 Swan bands are in any case better than sharp narrow atomic emission lines of Fe I, which for example would be produced by an iron nail! The lower electrode consists of the sample holder. For the fastening of different types and shapes of samples I've created a whole set of different versions. In the right picture a makeshift container for loose material or liquids can be seen (eg salt).


Who plans more extensive tests will probably better use a spark inductor. This may be con-structed for example using a simple car ignition coil and some additional electronics, which nowadays replaces the former terribly rattling mechanical interruptor. Under these key-words numerous construction guidances are to find on the Internet.

3.2 Design and Implementation of the Experiment

The DADOS spectrograph with CCD camera is mounted on a tripod. Absolutely necessary is a focus lens that produces a sharp image of the sample on the slit plate. For this purpose I use an ancient VIVITAR macro zoom lens. This requirement could also be met using a sim-ple close-up lens of appropriate focal length. The spectra are recorded here with a mono-chrome camera Meade DSI III Pro. The line identification was performed using the listed references, the “elements” or “lineident” tool of the Vspec software [52], as well as the NIST Atomic Spectra Database [31].

Logically, during the spark discharge the part of the generated plasma is recorded, which is located just on the slit of the spectrograph. Therefore, the impact point of the spark on the sample must be centered on the slit as precisely as possible. Preferred impact points are sharp tips, corners or edges.

The necessity of this procedure shows a simple exper-iment. If the upper graphite electrode is positioned on the slit, one obtains a spectrum of Swan emission bands with molecular C2, which is typical for carbon compounds (Table 1). Aimed on the center of the arc-over between the two electrodes re-sults in the characteristic spectrum of a lightning strike, with lines of ionised air compo-nents, ie above all, O II and N II (Table 2). Focused on the sample appears its emission spec-trum, possibly superimposed by the spark spectrum of the air molecules, because the test is carried out here without the protection of any argon or another inert gas. In this case, a previously recorded spark spectrum of the air must be subtracted or be displayed together with the sample spectrum. Thus, it becomes immediately clear which emissions originate


from the sample and which from the air molecules. Should only the spark spectrum of the air be visible, in most cases, the impact point of the spark was not positioned accurately enough on the slit of the spectrograph. Therefore, it is always worth to take several shots with a slightly altered perspective. The spark should hit diagonally from the side onto the sample and not run in the slit axis – such as it is shown on the picture above.

For the 200L grating in a darkened room, exposure times <15 seconds have been proven. During this time, the spark was triggered about 10 times. With the 900L grating about 30 seconds, combined with approximately 20 – 30 spark strikes are necessary. Here own ex-periments are essential. Who is tired of the noise from manually triggerd sparks should soon change to a sparc inducor. Computer monitors should be turned off during the record-ings or turned away because they produce itself relatively intense emission lines.

The following experiments were carried out with electrically conductive samples, which were used directly as electrodes. Non-conductive materials must first be prepared accord-ingly. In the literature, e.g. the coating of the surface with graphite powder, etc. is recom-mended. For liquids, [21] for example suggests the impregnation of coal pills, available in pharmacies.

3.3 Results

Table 1:

Table 1 shows the spectrum of the graphite electrode, i.e. the typical C2 Swan bands, pro-duced for example by comets and carbon stars [13]. Superimposed is here the spark spec-trum of the air molecules, which consists of emissions of ionised oxygen O II, -nitrogen N II and the hydrogen line Hα. The strong sodium emission (Fraunhofer D1 and D2) shows up in most such material spectra as a dominant byproduct or "contamination" due to the low ex-citation energy of Na. This was already noticed by Bunsen and Kirchhoff – excerpts from [20], translated here from German to English:

…Of all the spectral responses the one of sodium is most sensitive. The yellow line, the only one which sodium exhibits in the spectrum, coincides with the Fraunhofer D line and is characterized by a particularly sharp boundary and its exceptional brightness. If the flame temperature is very high and the amount of the applied substance very large a continuous spectrum in the immediate vicinity of the line becomes evident. Intrinsicly weak neigbouring lines of other substances appear to be even more weakened and some of them remain often invisible until the reaction of sodium expires…

…Already Swan has advised to the just small amounts of salt, which are necessary to gen-erate very intense sodium lines. The following experiment shows that in the field of chemis-try no other reaction can be compared, not even remotely, with the sensitivity of spectro-scopic determination of sodium. We deflagrated in the furthermost corner of the observa-tion room, which holds about 60 cubic meters of air, 3 milligrams sodium chlorate with milk sugar. After a few minutes a strong sodium line appeared, wich completely disappeared just after 10 minutes.

To a somewhat lesser extent than the sodium, also calcium lines (Ca I and Ca II) may ap-pear as a byproduct in such material spectra.


C

ompo

site

Spe

ctru

m:

Gra

phite

Ele

ctro

de+

Air

D

AD

OS

20

0L/

mm

©R

icha

rd W

alke

r 2

01

1/0

2

Lege

nd:

Bla

ck: E

mis

sion

line

sA

ir (N

II, O

II)

Red

: Em

issi

on li

nes

Gra

phit

eG

reen

: Byp

rodu

cts

or c

onta

min

atio

ns

Kom

posi

t G

raph

it +

Luf

t

N II 6482

Hα 6562.85

N II 6611

O I 7157

C2 5636C2 5585

C2 5541C2 5502

C2 5165C2 5130

N II 5004/6

C2 4737C2 4715

C2 4685

C2 4698

Na l 5890/96

Hα 6562.82

N II 5932/42

N II 5667/79

O II / N II

N II 4447

TABLE 1


Table 2:

Table 2 shows the spark spectrum of the air molecules, recorded in the middle between the two graphite electrodes. Here just emission lines of the air and no remnants of C2 Swan bands or any other carbon molecules can be detected.

This profile shows, not surprisingly, striking similarities with a lightning spectrum (see table 2a), which is dominated by the emissions of singly ionised nitrogen (N II) and oxygen (O II). The most intense lines are the cause for the bluish-white color of the lightning or spark. On-ly the usually strong hydrogen emissions, typically observed in the lightning spectra, appear here just weak in this experimental setup. Interestingly, however, they are more or less intensly visible in most material spectra, even if the assessed samples hardly contain any hydrogen.

Table 2a:

Table 2 shows a lightning spectrum, which was recorded via cloud reflection. The C8 tele-scope was set up in the living room at night with DADOS and Atik314L+, aiming through the closed window on the approaching thunderstorm. Three shots have been processed, each 180 seconds in 2x2 binning mode. Per exposure the reflected light of about 5-10 lightning strokes has been recorded. Since the cloud base was very low, on all shots the light pollution had to be subtracted. Striking is the intense CN emission at about 3900A.

Already since the beginning of spectroscopy in the 19 century, it was attempted to gain spectra of lightning discharges. To the beginning of the 20th century also well known as-tronomers have been involved like Pickering and Slipher. The lightning research today is particularly interested in the course of intensity, plotted against time, of the individual emission lines during a lightning strike. This allows conclusions about the physical pro-cesses that are active during the various phases of this phenomenon. The following dia-grams from [15], taken with high-speed cameras, show the intensity profile of a lightning spectrum in a time span of 0–50 μ sec and plotted in the range of 4,000-5,000 Å.


S

park

spe

ctru

mof

air

DA

DO

S 2

00

L/m

m

©R

icha

rd W

alke

r 2

01

1/0

2

Lege

nd:

Bla

ck:

Emis

sion

lin

esA

ir (

N II

, O II

)

Hα 6562.85

TABLE 2

N II 3995N II 4041

O II 4072/76N II 4097

O II 4119 - 53O II 4190

N II 4227 - 37

O II 4349

O IIN II 4447

N II 4510/14

N II 4630/43N II 4601/07

O II 4700/05

N II 4803.3N II 4780/88

Hβ 4861 / N II 4859

O II 4943

N II 5001 - 07 N II 5045

N II / O II 5179

N II 5535N II 5496

N II 5667/79

N II 5932/42


Li

ghtn

ing

Clo

udR

efle

ctio

nC

8, D

AD

OS

: Gra

ting

20

0L

mm

-1, S

litw

idth

25μ

m

4.8

.201

2A

tik

314

L+ -

2°C

, 3x1

80

s, 2

x2 B

inn

ing

(ca.

30

Lig

htni

ngst

roke

s)

©R

ich

ard

Wal

ker

20

12

/08

Hα 6562.83

N II 5001 - 07

N II/O II

Hβ 4861

N II / OI

N II 5171 -99

N I 6482 - 84

N II

N II 3995N II 4041 - 44

O II 6640.9O II 6721.35

N I 6008

O II 3910 - 26

Hδ /N II / O II

CN 3871 ?

N I 6441 - 57

N I 5560- 64N I 5616- 23

N I / N II / OI

Hγ / N I / O II

N II 5670- 90

N I

N I 4915

N I / N II

N I / O II

O I 6156- 58

O I 7157

O I

O I

TABLE 2a


Table 3 to 7:

The tables 3 to 7 show the spectrum of C15E steel/materialnumber 1.1141 (hereafter called “C15 E”). Chemical additives in [%] according to [62]: Chemische Zusammensetzung (in %) C carbon: 0.12 – 0.18 Si silicium: max. 0.40 Mn manganese: 0.30 – 0.60 P phosphor: max. 0.035 S sulfur: max. 0.035

Besides the main component Fe in the higher-resolution spectra only Mn (0.4%) is detecta-ble. The numerous Fe lines could theoretically be used to calibrate a spectrum. Since there are smarter ways however, as to fire up the spark coil each time. The profiles are unnormalised and uncorrected pseudocontinua.

Table 3: Low-resolution (200L) composite spectrum: C15E steel + spark spectrum of air.

Table 4: 900L composite spectrum in the range 3900 – 4800 Å of C15E steel + spark spectrum of air (blue profile). See also Table 20, where this spectrum is compared with the profile of the sun. Shown here is also the separately recorded air spectrum without iron lines (red curve).

Table 4a: C15E steel spectrum from Table 4 after subtraction of the air spectrum. Neverthe-less, here some weaker emission lines of the air can still be seen. The iron lines show up here more intensive (subtraction with the Vspec function "Subtract Profile by Profile").

Table 5: dito Table 4, here in the range of 4650 – 5550 Å

Table 6: dito Table 4, here in the range of 5550 – 6450 Å and without a superposed air spectrum, because almost no Fe I lines show up here.

Table 7: dito Table 6, here in the range of 6450 – 7400 Å

.


C

ompo

site

Spe

ctru

m:

C15

E St

eel +

Air

DA

DO

S 2

00

L/m

m

©R

icha

rd W

alke

r 2

01

1/0

2

Ca II 3933.66 Ca II 3968.47

Fe I 3878 Fe I 3856

N II 3995N II 4041 / Fe I 4046

Fe I 4071 O II 4119 - 53

O II 4190Ca I/N II/Fe I 4226 - 28

Fe I 4271Fe I 4308Fe I 4326

Fe I 4383.54Fe I 4404.75 N II 4447

N II 4601N II 4630/43

O II 4664/76O II 4706

Fe I 4871Fe I 4891

Fe I 4920Fe I 4958

N II 5001 - 07

Fe I 5139

Fe I 5227/33Fe I 5269-72

Fe I 5328Fe I 5369-71

Fe I 5406Fe I 5455

N II 5535

Fe I 5615N II 5667/79

Na l 5890/96

Hα 6562.82

K I 6964 ?

N II 4788

N II 5932/42

N II 4510/14

N II / O II 5179

TABLE 3Le

gend

:

Bla

ck: E

mis

sion

line

sA

ir (

N II

, O II

)R

ed: E

mis

sion

line

sst

eelC

15

E G

reen

: By-

prod

ucts

or c

onta

min

atio

ns


Com

posi

te S

pect

rum

: C

15E

Stee

l + A

irD

AD

OS

90

0L/

mm

©R

icha

rd W

alke

r 2

01

1/0

2

Ca II 3933.66

Ca II 3968.47

N II 3995

Fe I 4045.8

Fe I 4062

O II 4146

Ca I/Fe I 4227

Fe I 4271.15

Fe I 4307.9

Fe I 4325.76

Fe I 4383.55

N II 4447

N II 4601

N I 4630.54

O II 4662

O II 4700/05

Mn I 4057.9

Fe I 4071.7

O II 4076 N II 4097

O II 4119.22O II 4132.8

O II 4153

Fe I 4175O II/N II 4190/95

N II 4200

N II 4237Fe I 4250.8Fe I 4260.47

O II 4317.14

O II 4349

O II 4367-68

Fe I 4404.75O II/Fe I 4415

O II 4466

N II 4607N II 4614N II 4621

O II 4642 + N II 4643O II 4649

O II 4676

N II 4510/14

N II 4530

N II 4591

N II 4596

N II 4788

Com

posi

te s

pect

rum

C1

5 E

+ A

ir

Spa

rk S

pect

rum

Air

N II 4041

O II 4072/76

Fe I 4199

N II 4227Fe I 4236

O II 4396

TABLE 4

Fe I 4005

Lege

nd:

Bla

ck: E

mis

sion

line

sA

ir (

N II

, O II

)R

ed: E

mis

sion

line

sst

eelC

15

E G

reen

: By-

prod

ucts

or c

onta

min

atio

ns


C

15E-

Stee

l Spe

ctru

maf

ter

Sub

trac

tion

ofth

eA

ir S

pect

rum

90

0L/

mm

©R

icha

rd W

alke

r 2

01

1/0

2

Ca II 3933.66

Ca II 3968.47

Fe I 4045.8

Fe I 4062 Fe I 4071.7

Mn I 4057.9

Ca I/Fe I 4227

Fe I 4250.12Fe I 4260.47

Fe I 4271.15

Fe I 4307.9

Fe I 4325.76

Fe I 4383.55

Fe I 4404.75

TABLE 4a

N II 3995


©R

icha

rd W

alke

r 2

01

1/0

2

Fe I 4737

N II 4779.71

O II 4676.23

O II 4699.21O II 4705.36

N II 4788N II 4803.3

Hβ 4861 / N II 4859Fe I 4871

Fe I 4891

Fe I 4920.5

O II 4943Fe I 4957.6

N II 5001 - 5007

N II 5045

Fe I 5139

N II / O II 5179

Fe I 5075

Fe I 5192

Fe I 5227Fe I 5233

Fe I 5269 - 72

Fe I 5328Fe I 5324

Fe I 5341

Fe I 5369 - 71Fe I 5341

Fe I 5406

Fe I 5455

N II 5496

N II 5535

N II 5480

TABLE 5

Com

posi

te S

pect

rum

: C

15E

Stee

l + A

irD

AD

OS

90

0L/

mm

Com

posi

te s

pect

rum

C1

5 E

+ A

ir

Spa

rk S

pect

rum

Air

Lege

nd:

Bla

ck: E

mis

sion

line

sA

ir (

N II

, O II

)R

ed: E

mis

sion

line

sst

eelC

15

E G

reen

: By-

prod

ucts

or c

onta

min

atio

ns


©R

icha

rd W

alke

r 2

01

1/0

2

Fe I 5616

N II 5535

N II 5666.63N II 5679.56

N II 5710.77

Na I 5889.95Na I 5895.92

N II 5931.78N II 5941.65

TABLE 6

Com

posi

te S

pect

rum

: C

15E

Stee

l + A

irD

AD

OS

90

0L/

mm

Com

posi

te s

pect

rum

C1

5 E

+ A

irLe

gend

:

Bla

ck: E

mis

sion

line

sA

ir (N

II, O

II)

Red

: Em

issi

on li

nes

stee

lC1

5E

Gre

en: B

y-pr

oduc

ts o

r con

tam

inat

ions


©R

icha

rd W

alke

r 2

01

1/0

2

N II 6482

Hα 6562.85

N II 6611

O I 7157

TABLE 7

Com

posi

te S

pect

rum

: C

15E

Stee

l + A

irD

AD

OS

90

0L/

mm

Com

posi

te s

pect

rum

C1

5 E

+ A

ir

Lege

nd:

Bla

ck: E

mis

sion

line

sA

ir (

N II

, O II

)R

ed: E

mis

sion

line

sst

eelC

15

E G

reen

: By-

prod

ucts

or c

onta

min

atio

ns


Table 8: Low-resolution (200L) composite spectrum: Galvanized steel + spark spectrum of air.

Striking are here, as expected, the emission lines of zinc. “Front runner” is clearly the line at 4810.53 Å. In this profile it is clearly noticeable that the emission lines of the air show up relatively modest, particularly the otherwise intense line, N II 5001 – 5007 Å, is missing here even totally. There is also no strong continuum visible. The reason for this effect re-mains unclear until now. Perhaps the spectrograph slit was accidentally aligned here very precisely to the impact site of the spark.

Table 9: Low-resolution (200L) composite spectrum: aluminum + spark spectrum of air.

Here three aluminum emissions in the blue region of the spectrum are dominant. Also in this profile, it is striking that the emission lines of the air appear relatively modest and a strong continuum is missing. Two clusters with emission lines in the range of 4600 – 4750 Å and 5100 – 5200 Å, can currently not be clearly identified.

Table 10: Low-resolution (200L) composite spectrum: copper wire + spectrum of air.

At first the insulation varnish was removed from the copper wire. Also here a strong contin-uum is missing. Striking are here three dominant copper lines in the range of 5100 - 5300 Å. Several others are blended with N II and O II emissions, which appear here relatively in-tensive. Noteworthy are also the by-products, i.e. Na I and Ca II lines.

Table 11: Low-resolution (200L) composite spectrum, recorded with Urs Flükiger in his company in Oberburg, Switzerland. Sample: Tool steel coated with titanium oxide (Ti O2).

Well visible are here emission bands of diatomic titanium monox-ide molecules (TiO) and atomic Ti lines. Titanium dioxide (Ti O2) at such high temperatures > 3000 K, dissociates to titanium monox-ide (Ti O) like in the stellar atmospheres, down to late M-Class.

The air spectrum appears here only weak. See also Table 21, where a section of this spectrum is shown in a montage with the profile of Ras Algethi (α Her). In the area of 5000 Å, however, the leading line of the air spectrum, the intense emission of N II, is blended with the TiO bands.

Because in this experiment the spark had to penetrate the titanium oxide layer, it was cor-respondingly weaker. Therefore to improve the recognisability of the absorption bands, the exposure time was increased here to 1 minute and the spark was triggered approximately 150 times. If the spark hits an edge or corner of the sample, there is a risk that the coating is here too thin or no longer present. Therefore it has proven to be ideal, when the spark hits the center of the sample at a right angle and the picture is recorded approximately in the surface plane. So it does not matter if the impact point slightly moves.

Remark: Several attempts failed so far, to generate emission lines of gold. In each case only the spectrum of the air molecules was obtained. The reason is still unknown.


©R

ich

ard

Wal

ker

20

11

/02

N II 3995

N II 4041

N II 4195

Na l 5890/96

TABLE 8

Zn I 4680.14Zn I 4722.15

Zn I 4810.53

Zn I 6362.34

Mn I ?

O II/N II ?

Mn I ?

N II 4097Mn I ?

Com

posi

te S

pect

rum

: G

alva

nise

dSt

eel +

Air

DA

DO

S 2

00

L/m

m

Lege

nd:

Bla

ck: E

mis

sion

line

sA

ir (N

II, O

II)

Red

: Em

issi

on li

nes

galv

anis

edst

eel

Gre

en: B

y-pr

oduc

ts o

r con

tam

inat

ions


©R

icha

rd W

alke

r 2

01

1/0

2

Na l 5890/96

TABLE 9

Al I 3961.52

Hα 6562.85

Al I 3944.0

Al II 4226.81

N II 5667/79

N II / Hβ 4859/61

O II ?

C2 ?

N II 5001 - 07

Com

posi

te S

pect

rum

: A

lum

inum

+ A

ir

DA

DO

S 2

00

L/m

m

Lege

nd:

Bla

ck: E

mis

sion

line

sA

ir (

N II

, O II

)R

ed: E

mis

sion

Alu

min

umG

reen

: By-

prod

ucts

or c

onta

min

atio

ns


©R

icha

rd W

alke

r 2

01

1/0

2

Na l 5890/96

TABLE 10

Hα 6562.85

Cu I 5218.2

N II 5667/79

Cu I 5153.24Cu I 5105.54

Ca II 3933.66Ca II 3968.47

Cu I 4062.6 / O II 4349

Cu I 4651.12 / N II 4630/43

Cu I / N II 5001 - 07

N II 3995N II 4041

O II 4119 - 53

O II 4190Cu I 4230 / N II 4227 – 37 / Ca I

Cu I 4275.1

N II 4447

O II 4700/05

N II 4803.3

Cu I 5292.52

Cu I 5700.24

Cu I 5782.13

N II 5932/42

Cu II 6154.22

Fe I 3906.48 ?Com

posi

te S

pect

rum

: C

oppe

r W

ire

+ A

ir

DA

DO

S 2

00

L/m

m

Lege

nd:

Bla

ck: E

mis

sion

line

sA

ir (

N II

, O II

)R

ed: E

mis

sion

Cop

per

Wir

eG

reen

: By-

prod

ucts

or c

onta

min

atio

ns


©R

icha

rd W

alke

r 2

01

1/0

3

Ti I 3989.76

TABLE 11

Ti I 4055

Ca II/Ti 3934

Na I 5889/95

TiO 5497

TiO 5448/51

TiO 4761-63 TiO 4704

TiO 4626 TiO 4668

TiO 4804

TiO 5167

TiO 5003/5020

TiO 6268

Hα 6562.82

N II 5001 - 07

Ti I 4137O II 4190

Ti I 4263

Ti I 4300Ti I 4390/93

N II 4447

Ti I/ TiO 4531/36

?

?

?

N II / O II 5179

Com

posi

te S

pect

rum

: Tit

aniu

m M

onox

ide

(Ti O

) Coa

ting

+ A

ir

DA

DO

S 2

00

L/m

mLe

gend

:

Bla

ck: E

mis

sion

line

sA

ir (N

II, O

II)

Red

: Em

issi

on T

itan

ium

Mon

oxid

eG

reen

: By-

prod

ucts

or c

onta

min

atio

ns


4 Tests with Arc Excitation

4.1 Design and Implementation of the Experiment

Professionally arc spectra are obtained in an inert protective gas atmosphere with elec-trodes, made of tungsten or spectrally pure high density graphite. In the factory of Urs Flükiger in Oberburg Switzerland [61] we have attempted to simulate this rudimentary with an ordinary electric welding machine and a rutile-acide type electrode (Ti O2). An additional test was carried out with a TIG electrode, combined with argon gas (TIG: Tungsten-Inert-Gas).

Picture: Electric welding with rutile coated electrode. DADOS spektrograph and old VIVITAR macro zoom lens on an AYO mount, distance ca. 3m. Urs Flückiger (left picture in the back-ground), recording the spectra on the laptop.

TIG welding with Argon protective gas (Mr. Baumgartner)


4.2 Results

Table 15: Electric arc, simulated with electric welding machine and rutile-acide type elec-trode (Ti O2). Low-resolution (200L) composite spectrum: steel 37, materialnumber 1.0037, (herafter called “steel 37”) + spectrum of air. The comparison with the spark spec-trum shows that the arc spectrum is resolved somewhat higher. In addition to the lines of Fe I atoms and air molecules here very numerous atomic and molecular emissions of titani-um appear (Ti / TiO), obviously originating from the mineral rutile (Ti O2) of the welding electrode. The distinction between Fe I and Ti I has proven to be difficult and uncertain. These lines are often separated only by a fraction of 1Å. Also striking is the broad Na I emission with a central self-absorption dip, as can be observed in the spectra of high-pressure sodium vapor lamps (see [13]).

Table 16: Electric arc, simulated with electric welding machine, TIG electrode, combined with argon gas. Low-resolution (200L) composite spectrum: steel 37 + spectrum of air.

In older welding systems radioactive thoriumoxide improved the ignition. This tungsten electrode is free of radioactive thorium here because Flükiger & Co Ltd applies modern welding equipment and high-voltage ignition with appropriate control system. Consequent-ly, such emission lines are her absent in the spectrum.

The profile is impressively dominated by intensive emissions of argon protective gas Ar I. To make the Fe I lines at all visible, a strong zoom in to the intensity axis was necessary. In the red region of the spectrum, therefore most of the Ar I emission appear cutted. In reality, they are about two to three times as high! Clearly visible here is the protective effect of ar-gon gas. The emissions of the air molecules are weak. The "front runner line" N II at 5000 – 5007 Å is here even completely missing!


C

ompa

riso

nA

rcS

pect

rum

Stee

l 37

ver

sus

Spa

rk S

pect

rum

C15

E

DA

DO

S 2

00

L/m

m

©R

icha

rd W

alke

r 2

01

1/0

3

Ca II 3933.66 Ca II 3968.47

Fe I 3878

N II 3995N II 4041 / Fe I 4046/71

O II 4119 - 53O II 4190

Ca I/N II/Fe I 4226 - 28

Ti I 4300/Fe I 4271 4326

Fe I 4383.54

N II/OII

Fe I 4871Fe I 4891

Fe I 4920Fe I 4958

N II 5001 - 07

Fe I 5170

Fe I 5227/33Fe I 5269-72

Fe I 5328Fe I 5369-71

Fe I 5406Fe I 5455

N II 5535

N II 5667/79

Na l 5890/96

Hα 6562.82

N II 5932/42

N II / O II 5179

TABLE 15

Fe I 4405

Ti I 4531

Ti I 4457

TiO 4761/63TiO 4761/63

Ti I 48251

Ti I 5038

Fe I 4985-87

Fe I 5065

Fe I 5205

Fe I 5512-13

Fe I 5800

Fe I 6015-18

TiO 6159

TiO 6268

Fe I 6710 – 13 ?

Fe I 6917 ?Fe I 6945 ?

Spa

rk S

pect

rum

C1

5E

Arc

Spe

ctru

mSt

ahl 3

7

Lege

nd:

Bla

ck: E

mis

sion

line

sA

ir (

N II

, O II

)R

ed: E

mis

sion

ste

el3

7/C

15

EG

reen

: By-

prod

ucts

or c

onta

min

atio

ns


©R

icha

rd W

alke

r 2

01

1/0

2

Com

posi

te A

rcS

pect

rum

: S

teel

37

+ A

ir

DA

DO

S 2

00

L/m

m

Ca II 3933.66 Ca II 3968.47

Fe I 3878 Fe I 3856

Ar I 4158 / 64

Fe I 4891Fe I 4920Fe I 4958

Fe I 5227/33Fe I 5269-72Fe I 5328Fe I 5369-71

Fe I 5455

Na l 5890/96 ?

Hα 6562.82

TABLE 16

N II 4041 / Fe I 4046/71

Ar I 4198 / 00

Ar I 6965.43

Ar I 7158.84

Ar I 7272.94

Ar I 7384

Ar I 7067.22

Ar I 6752.83

Ar I 4259 / 66 /72Ar I 4300

Ar I 4333 / 35 / 45

Ar I 4510 / 22

Ar I 4628.44Ar I 4596

Ar I 5650

Ar I 5496

Fe I 5914

N II / O II 5179

Fe I 5406

Fe I 5563 - 66Fe I 5610

Fe I 5740Fe I 5835

Fe I 5860/61

?Fe I 6220

?

?

Fe I 6416 - 19

N II 4780/88

Fe I 4383.54N II 4447

Ar I 6871.3Ar I 6937.66

Ar I 7030.25

Ar I 7206.98

Ar I 7316

?

?

Lege

nd:

Bla

ck: E

mis

sion

line

sA

ir (

N II

, O II

)R

ed: E

mis

sion

ste

el3

7G

reen

: By-

prod

ucts

or c

onta

min

atio

nsB

lue:

Em

issi

on li

nes

prot

ecti

vega

s A

r


5 The Link between the Stellar- and Laboratory Spectra Table 20:

Shows in the blue range the comparison of the higher resolution (900L) composite spec-trum of C15E (according to Table 4) and the solar spectrum. All Fe I and Ca II lines of the sparc spectrum, can be clearly identified in the solar spectrum. Several iron lines in the spark spectrum are overprinted here by the much more intensive emissions of the air mole-cules (O II, N II) and therefore not available for a link to the solar spectrum.

Table 21:

Profile comparison, range from 4650 - 5900 Å, titanium oxide bands (TiO) (200L grating):

– Spark spectrum TiO emission bands of a surface coating (according to Table 11)

versus

– Stellar TiO absorption bands of Ras Algethi (α Her).

In this wavelength range in the spectrum of the surface coating appear clearly visible emis-sions of TiO bands. The intensity runs approximately mirror inverted to the characteristic absorption bands of the late M-class star α Her. The emissions of air molecules are here relatively weak. At 5000 Å however, N II forms an intense blend with the TiO bands.


Com

pari

son

ofS

pect

ra:

Sun

ver

sus

C15

E (+

Air

) 9

00

L/m

m

©R

icha

rd W

alke

r 20

11

/02

Ca II 3933.66

Ca II 3968.47

N II 3995

Fe I 4046

Fe I 4062

O II 4146

Fe I 4227 / Ca I

Fe I 4271

Fe I 4308

Fe I 4326

Fe I 4383

Fe I/Mn I 4058

Fe I 4072O II 4076

N II 4097

O II 4119.22

O II 4132.8

O II 4153

Fe I 4175

O II/N II 4190/95N II 4200

Fe I 4250Fe I 4260

O II 4317.14

O II 4349

O II 4367-68

Fe I 4405Fe I 4415

Com

posi

te s

pect

rum

C1

5 E

+ A

ir

Spe

ctru

mof

the

Sun

N II 4041

O II 4072/76

Fe I 4199

N II 4227

Fe I 4236

Fe I 4005

CH 4299 - 4313

Hγ 4340.47

Hδ 4101.74

TABLE 20Le

gend

:

Bla

ck: E

mis

sion

line

sA

ir (

N II

, O II

)R

ed: E

mis

sion

ste

elC

15

EG

reen

: By-

prod

ucts

or c

onta

min

atio

nsB

lue:

Abs

orpt

ion

lines

spec

trum

ofth

esu

n


C

ompa

riso

nTi

O2

Coa

ting

and

Ras

Alg

ethi

(αH

er M

5 Ib

-Il)

20

0L/

mm ©

Ric

hard

Wal

ker

20

11

/03

N II 4859

Com

posi

te s

pect

rum

TiO

2C

oati

ng+

Air

Spec

trum

Ras

Alg

ethi

TABLE 21TiO 5847

TiO 5597/03

TiO 5497

TiO 5448/51

TiO 5359

TiO 5308

TiO 5240

N II 5001 - 07TiO 4955

TiO 4847

TiO 4761

TiO 4704

TiO 4668

TiO 4804

TiO 5167

TiO 5814

TiO 5668

TiO 5760

Na l 5889/95

N II / O II 5179

TiO 5003/20 5020


6 Appendix

6.1 Kirchhoff's hand-drawn Spectra [20].

The Fraunhofer absorption lines of the sun (top line) compared to the emission spectra of the elements Ka (today K), Na, Li, Sr, Ca und Ba.


6.2 Vatican Spectral Atlases of Elements and Molecules

The Vatican Observatory in 1930-40 published several fundamental works with spark and arc spectra, continuing the tradition of the legendary Father Secchi, who wrote in the 19th century the first stellar spectroscopic atlas. The driving force behind this project was Father Alois Gatterer. The following pictures are from the antiquarian book J. Ricerche Spettroscopiche, from J. Junkes and E.W. Saltpeter [1].


7 Literature and Internet

Literature:

[1] Ricerche Spettroscopiche, J. Junkes, E.W. Salpeter, 1966, Laboratorio Astrofisico Della Specola Vaticana. Dreissig Jahre Spektralatlanten des Astrophysikalischen Laboratoriums der Vatikanischen Sternwarte. Vergriffen, nur noch antiquarisch erhältlich.

Internet Links:

Author:

The following scripts on the subject are downloadable under this link (some of them in german): http://www.ursusmajor.ch/astrospektroskopie/richard-walkers-page/index.html

[10] Analysis and Interpretation of Astronomical Spectra, Theoretical Background and Practical Applications for Amateur Astronomers

[11] Das Aufbereiten und Auswerten von Spektralprofilen mit den wichtigsten IRIS und Vspec Funktionen

[12] Kalibrierung von Spektren mit der Xenon Stroboskoplampe

[12a] Kalibrierung von Spektren mit dem Glimmstarter ST 111 von OSRAM

[13] Spectroscopic Atlas for Amateur Astronomers

[14] Quasar 3C273, Optical Spectrum and Determination of the Redshift

Lightning Spectroscopy:

[15] A High-Speed Time-Resolved Spectroscopic Study of the Lightning Return Stroke: Part I, a Quali-tative Analysis, Richard E. Orville, Institute of Atmospheric Physics, Tucson Arizona. http://journals.ametsoc.org/doi/pdf/10.1175/1520-0469(1968)025%3C0827%3AAHSTRS%3E2.0.CO%3B2

[16] The optical spectrum of aircraft St. Elmos fire, E.M. Wescot et al. University of Alasca, 1996. http://www.uas.alaska.edu/artssciences/naturalsciences/envs/faculty_staff/pubs/wescott96_stelmo.pdf

[17] Lightning Physics and Effects, V. Rakov, M. Uman

[18] The Spectrum of Lightning, L. Wallace, Kit Peak, National Observatory http://adsabs.harvard.edu/full/1964ApJ...139..994W

[19] Lightning Spectroscopy, T. Walker, H. Christian, D. Sentman

Emission Spectroscopy (German language)

[20] Chemische Analyse durch Spectralbeobachtungen, G. Kirchhoff R. Bunsen, Annalen der Physik und [20]Chemie 1860, http://www.ub.uni-heidelberg.de/helios/fachinfo/www/physik/erg/spek.pdf

[21] http://www.students.uni-mainz.de/jungc000/uugit/inhaltsv.htm Inhaltsverzeichnis mit zahlreichen Ideen und Tips zur Funken- und Flammenspektroskopie u.a. unter der Rubrik „Chemische Analysen“.

[22] Phasenaufgelöste, optische Emissionsspektroskopie an RF Plasmarandschichten, Diplomarbeit 2007, S. Nemschokmichal, Universität Greifswald http://www1.physik.uni-greifswald.de/download/dissertationen/dip-nemschokmichal.pdf

Database

[31] NIST Atomic Spectra Database: http://physics.nist.gov/PhysRefData/ASD/lines_form.html

[32] Simulated emissionspectra of all elements, http://bmauclaire.free.fr/astronomie/spectro/simulation/noblet/index2.htm


Metallurgy and Welding Technology (German language)

[61] Flükiger & Co AG - Industrieschmiede, Emmentalstrasse 75, CH 3414 Oberburg http://www.fluekiger.ch/

[62] Werkstoffdatenblatt Voegelin AG, C15E Stahl/Werkstoffnummer 1.1141 http://www.voegelinag.ch/DE/Dokumente/Werkstoffdatenblatt_C15E.pdf

Documents

Atomic Emission Spectroscopy with Spark- or Arc · PDF fileAtomic Emission Spectroscopy with Spark or Arc Excitation 4 2 Excitation of Emissions Spectra Emission lines can not only