Analysis of the conversion coating on ferritic stainless steel of selective absorbers

Analysis of the conversion coating on ferritic stainless steel of selective absorbers

Solar Energy Materials 14 (1986) 143-159 North-Holland, Amsterdam 143 ANALYSIS OF THE CONVERSION COATING ON FERRITIC STAINLESS STEEL OF SELECTIVE AB...

1MB Sizes 0 Downloads 22 Views

Solar Energy Materials 14 (1986) 143-159 North-Holland, Amsterdam

143

ANALYSIS OF THE CONVERSION COATING ON FERRITIC STAINLESS STEEL OF SELECTIVE ABSORBERS L. A R I E S , P. F O R T , J.A. F L O R E S a n d J.P. T R A V E R S E * Laboratoire de Recherche sur I'Energie, Universitb Paul Sabatier, 31062 Toulouse Cedex, France

Received 21 March 1986; in revised form 3 June 1986 The surface characterisation of selective solar absorbers prepared with an original conversion treatment of ferritic steel is achieved with different methods: microscopy (SEM and high voltage electron microscopy), secondary ion mass spectroscopy (SIMS), electron spectroscopy for chemical analysis (XPS). This study brought to light the complex nature of the coating composed of 3 zones. In varying proportions, according to the depth, the main components are iron oxides (basically magnetite) and hydroxides which are probably chromium substituted and particles of unoxidised metal with small quantities of sulphur-containing compounds.

1. Introduction F o r m a n y years n u m e r o u s studies have a i m e d at d e v e l o p i n g selective a b s o r b e r s for p h o t o t h e r m i c solar energy conversion. These surfaces are characterized b y a high solar a b s o r p t i v i t y a n d a low emissivity ( a s in the region of 0.95 a n d c~ a r o u n d 0.20). These p r o p e r t i e s can be o b t a i n e d either b y m o d i f y i n g the state of the surface or, m o r e generally, b y a p p l y i n g a d e p o s i t on sheet-metal [1-6]. W e d e v e l o p e d two processes for the p r e p a r a t i o n of selective surfaces from stainless steels. T h e y involve either a n o d i c o x i d a t i o n or chemical t r e a t m e n t [7,8]. T h e c o m p o s i t i o n of the conversion coatings is c o m p l e x a n d d e p e n d s on the n a t u r e o f the s u b s t r a t e a n d on the p r e p a r a t i o n c o n d i t i o n s [9-12]. The present w o r k focuses on the c h a r a c t e r i s a t i o n of this type of selective a b s o r b e r in o r d e r to explain the p a r t i c u l a r optical p r o p e r t i e s in relation with the c o m p o s i t i o n , texture a n d state of crystallization.

2. Experimental 2.1. Selective surfaces preparation and experimental setup

T h e s t u d y was carried out with c h e m i c a l l y - t r e a t e d a b s o r b e r s which were o b t a i n e d from ferritic stainless steel c o n t a i n i n g n i o b i u m ( F r e n c h S t a n d a r d Z8 C N ~ 1 7 ) whose * Member ISES. 0 1 6 5 - 1 6 3 3 / 8 6 / $ 0 3 . 5 0 © Elsevier Science Publishers B.V. ( N o r t h - H o l l a n d Physics Publishing Division)

144

L. Aries et al. / Analysis of the coating on ferritic steel

Table 1 Chemical composition of the ferritic stainless steels % Weight

c Cr Ni Mo Cu Si S Ti Nb P A1 Sn Mn Fe

Ferritic sheet (Z8CNbl7)

Ferritic foil

0.035 16.75 0.24 0.02 0.05 0.275 0.014 0 0.76 0.024 0.056 0 0.53 81.25

0.080 19.75 0.18 0.01 0.05 0.240 0.015 0 0 0.030 0.074 0.105 0.54 78.93

composition is given in table 1. This cheap steel is c o m m o n l y used to m a n u f a c t u r e solar energy panels because of its good corrosion resistance in various media a n d because it is easy to work (e.g. cold folding, welding). The experimental setup was composed of a thermostated cell, a reference electrode (saturated calomel electrode) a n d an electronic millivoltmeter allowing the sample potential to be measured during the treatment. Before surface conversion, the sheet was cleaned with ethanol then rinsed with demineralized water. After the chemical treatment the sample was rinsed with demineralized water then oven-dried at 9 0 ° C . It should be noted that this process presents the advantage of only requiring one bath. Two bath compositions were c o m m o n l y used (table 2), at 45 o C. Bath B gave higher emissivity b u t emissivity is improved after thermal treatment. After 10 m i n treatment the surface o b t a i n e d in A was characterized by a~ = 0.95 and E~ = 0.20 a n d in B, after 7 min, by a s = 0.95 a n d ~~ = 0.25. These coatings are called " t y p i c a l selective coatings".

Table 2 Chemical compositions of the chemical treatment baths

Bath A Bath B

Sulfuric acid H 2SO4 (% vol.)

Propargyl alcohol H-C ~- C - CH :OH (mol/1)

Hydrated sodium thiosulphate Na2S203, 5H20 (tool/l)

5 15

0.10 0.12

2.0×10 3 2.0×10 3

L~ Aries et al. / Analysis of the coating on ferritic steel

145

2.2. Characterization methods Measurement of optical properties We measured the total near-normal solar absorptance a s (ratio between the energy absorbed by the collecting surface and the incident solar energy) with a reflectometer (Elan Informatique EL 510). The total hemispherical emittance c= (ratio between the energy emitted by the collecting surface and the energy emitted by a black body at the same temperature) was measured from absorption measurements of radiations emitted by a radiator at 7 0 ° C (IR absorptiometer Elan Informatique EL 520).

Secondary ion mass spectroscopy (SIMS) This technique allows the distribution profiles of the metallic and metalloidic elements in the coating to be obtained. Two operating methods were used. With oxygen in the chamber, metal signals are intensified and the proportion of ionized species is comparable in oxide or metal - the analysis is therefore semiquantitative. On the other hand if the analysis is done without any oxygen the ionization of a metallic element depends on its oxidation state: so results can supply information on the distribution of the compounds of a given element. The analysed zone diameter was approximately 100 # m wide and its depth is less than 2 nm.

Electron spectroscopy for chemical analysis (ESCA / XPS) The ESCA analysis depth is in the region of 5 to 10 nm. The thickness of a typical conversion coating is from 100 to 200 nm. Our process consists of successive scattering by means of argon ion bombardment. Thus the coating was analysed at different depths and we obtained the X-ray photo-electron spectra for various elements with excitation by A1 K~ radiation.

High-voltage electron microscopy and diffraction The high-voltage electron microscope used was that of the "Laboratoire d'optique 61ectronique" (CNRS-Toulouse). This type of microscope presents the advantage of having a great penetration power and a high resolution limit. We examined treated thin foil which we obtained from a 0.2 m m thick stainless steel sheet by electrolytic thinning and chemical conversion treatment.

3. Optical parameters The chemical treatment of the ferritic stainless steel containing niobium gives a remarkable selective solar absorber for photothermic solar energy conversion: a s >~ 0.95 and c~ ~< 0.20. This treatment is described in the following references [7,8,121. The ferritic steel chemical conversion coatings show an excellent stability in water and in humid air. A twelve week test did not cause any alteration of the appearance or of the optical properties of the coating. In the presence of chloride ions, the resistance to pitting is likewise comparable to that of the initial stainless steel [11,12]. In normal, on site working conditions the absorbers show no corrosion after 4 operating years (test in urban area).

146

L. Aries et al. / Analysis of the coating on ferritic steel

1.0 oo

0.9

B

0.8

m

~s



0.3 0.2 0.1 I

0

100

Exposure temperature I I I

200

300

400

(C) I

,

500

Fig. 1. Optical properties of the selective surfaces versus temperature of the thermal treatment (time of treatment 384 h).

We studied the influence of various thermal treatments in the air at temperatures between 20 and 400°C for a period of 16 d (384 h). Fig. 1 gives a, and ~ against temperature treatment period of 384 h. We may distinguish 2 temperature fields (fig. 1): - from 20 to 150°C the optical properties do not change; - f r o m 150 to 450°C E~ decreases and then becomes stable, a~ remains fairly constant. The selectivity of the coating improves.

4. Morphology of the coating

4.1. Macroscopic observation Visually, the conversion coating presents a homogeneous satiny black aspect. After the withdrawal of a superficial and very thin film which covers the layer by means of adhesive tape, the coating appears mat black. Nevertheless the optical properties remain almost unchanged after this operation (a~ >/0.95, c ~ ~<0.20).

4.2. Microscopic observation Examination by scanning electron microscopy (SEM) shows that the state of the surface of the material is quite homogeneous and has a slight roughness (fig. 2). After removal of the superficial film, the coating appears more clearly. Fig. 3 shows

L. Aries et al. / Analysis of the coating on ferritic steel

147

Fig. 2. SEM micrograph of the typical selective coating before the removal of the superficial thin film (a~ = 0.95, c ~ = 0.20).

that the surface of the material is regularly covered by pores or cavities of the order of 1 ~m, g r a i n - b o u n d a r i e s also seem to be hollowed, Crystals with geometrical forms measuring a few # m are characteristic of carbides. The grains of the steel have a m e d i u m size between 10 and 1 5 / t m a n d the g r a i n - b o u n d a r i e s measure less than 0.1 /.tm in width. Heat t r e a t m e n t widens these hollowed grain-boundaries. The evaluation of the surface roughness by m e a n s of alphastep profilometry reveals that

Fig. 3. SEM micrograph of the typical selective coating after the removal of the superficial thin film (a S= 0.95, c~ = 0.20).

148

L. Aries et a L / A nalysis of the coating on ferritic steel

Fig. 4. High voltage TEM micrograph (bright field image) of a specimen chemically treated for 7 min.

it may reach 200 nm in regions corresponding to the grain boundaries but is usually below 50 nm on the grains. Observation with a transmission electron microscope carried out on an electrolytically thinned foil treated in typical conditions allowed micrography of the material at very high magnification (fig. 4). It was then possible to see the material at greater magnification: the surface is covered with protuberances of some tenths of a micrometers. They look like certain types of crystallization in hillock. Between them, cavities appear and the deepness is in the region of some tenths of a micrometer, although it is difficult to evaluate exactly.

5. Chemical composition of the conversion coating 5.1. Secondary ion mass spectroscopy (SIMS) This process is based on ion sputtering of a zone of the material that is to be analysed: the sputtering time is related to the depth. But we have to bear in mind that the sputtering rate may vary according to the nature of the material. Semi-quantitative analysis of the coating: profiles of the main elements (analysis under oxygen) (figs. 5 and 6). We studied the ion intensity of different secondary ions of metals (Fe ÷, Cr ÷, Ni ÷) and metalloids ( O - , S-, C ) according to the sputtering time (figs. 5 and 6). The profiles obtained bring to light several zones. - a superficial zone which corresponds to a very short sputtering time (100 s) for which the ionic intensity of all the elements is extremely high.

L. A ties et al. / Analysis of the coating on ferritic steel /_superficial zone • i deep zone ,

r"

steel

","

Intensity

40.103

149

(cts.s 1) i I I

30.103

20.103 10.10 3 -

0 0

Ni,+x 2 ..................... I I '1 600 1200 1800

.. I 2400

= 3000

Time (s L . 3600

Fig. 5. SIMS intensity versus sputtering time for Cr +, Fe +, and Ni + for a typical selective coating.

a deep zone, where the iron and the chromium content of the coating increases closer to the metallic substrate, whereas the oxygen, sulphur and carbon contents decrease. After removal of the superficial film with adhesive tape profiles are obtained that are identical to those in the deep zone. We believe that there is a difference in the cohesion or compactness between the superficial zone and the deep zone. It may induce variations in the ionic sputtering rate which could explain the rise in the profiles in the superficial zone (figs. 5 and 6). Indeed, the ionic emission is higher if the coating removed in a given period of time is thicker. Identification of the main compounds of the coating (analysis without oxygen, figs. 7, 8, 9, 10, 11, 12). According to Namdar-Irani [13] and Alexandre et al. [14] the relative intensity of the secondary Fe30 ÷ (intensity versus Fe ÷ ion intensity i.e. I(Fe30 ÷ ) / I ( F e +) varies according to the nature of the emitting iron oxide. This relative intensity increases when we go from Fe203 through Fe304 to FeO. -

Intensity

100.10 3

(cts.s J)

75.10 3

~

50.10 3 25.10 3 0 0

300

x 1

.

I

600

900

"

1200

.I

1500

Timq ('s)~ 1800

Fig. 6. SIMS intensity versus sputtering time for O - , C - , and S - for a typical selective coating (sputtering rate is different to that in fig. 5).

150

L. Aries et aL / Analysis of the coating on ferritic steel I n t e n s i t y (cts.s 1)

10,104 8.104

6 .104

~ ' ~ ~ Cr.._*' 1...L. ,

4 .104

Fe+~4. i

2.104

~e30~'x 30

• ~,

0 0

. . . . ~*'lf- . . I NI xl01 600 1200 1800

I 2400

, Time ( s ) 3000 3600

Fig. 7. S I M S i n t e n s i t y versus s p u t t e r i n g time for F e 3 0 +, Fe +, C r + a n d Ni ÷ for a typical selective c o a t i n g . P r e p a r a t i o n time: 10 m i n , a , = 0.95, c ~ = 0.20.

5.104

I n t e n s i t y (cts.s .1)

4,104

3. 104 2.104 ~. C2xl ! ~ / O H - x l

1.104

0 0

600

1200

1800 2400 3000

3600 "

Fig. 8. S I M S i n t e n s i t y versus s p u t t e r i n g t i m e f o r O - , C2-, O H - , a n d S - for a typical selective c o a t i n g . P r e p a r a t i o n time: 10 m i n , a S = 0.95, c = = 0.20.

Intensity ( c t s s 1)

1 0 104 8 .104 6'104

,~.~.......- ~ ~ ,

4 104 .

~

C~r~ 1 Fe_O x30"" ~.....,.

2.104

-""~--...... :'-

0

F~+~A

Ni+xl0



I . . . . 1 . . . . . . .L__ _ • 1200 2400 3600 4800

0

i 6000

Time ( s ) .... 7200

Fig. 9. S I M S i n t e n s i t y versus s p u t t e r i n g time for F e 3 0 +, Fe +, C r + a n d Ni + for a thick c o a t i n g . P r e p a r a t i o n time: 60 rain, a , = 0.96, c ~ = 0.71. 5.104

ilntensity

(cts.s~)

4.104 3.104 2,104 1.104

0

~/'~.~/ % \~

"" "~ . . . . . .

9£5 J. _

" " "-~'~..... OH'x1 .S" x 10 ~. . . . . . . L.--~-_------I~_~' n I Time {s)_ 0 600 1200 1800 2400 3000 3600

Fig. 10. S I M S i n t e n s i t y versus s p u t t e r i n g time f o r O - , C 2 , O H time: 60 m i n , a~ = 0.96, ~ ~ = 0.71.

a n d S - for a thick c o a t i n g . P r e p a r a t i o n

L. Aries et al. / Analysis of the coating on ferritic steel superficial f i l m / e x t e r n a l zone // internal - • t!tJ.t, zone t steel Intenstyll (CtS.S-1) 10.104 8.10

[

104



5 . 1 0 -2

I (Fe +)

/

4 . 1 0 .2

' V /~'~.~__

4r 104

6.10 2

(Fe30.~'

/

4

6 . 1 0 4 /'~'% /

2

I (Fe30*l~( Fe+)

I

~ 1

.........

°'°~'1 ~'"

151

3 . 1 0 -2 Fe"× 4

2 . 1 0 .2 1 . 10 .2

Fe30,x30

I

I

I

I



6 0 0 12(X)1800 2 4 0 0 3OGOTime s) 0

0 0

Fig. 11. SIMS intensity of Fe ÷ and F e 3 0 ÷ and relative intensity l ( F e 3 0 + ) / l ( F e + ) versus sputtering time for a typical selective coating.

The different profiles obtained (figs. 7, 8, 11) show that the conversion coating is composed of several strata which are probably characterized by different iron oxides. The different zones of fig. 7 can be less clearly defined depending on the spot analysed. The layer prepared in typical conditions (preparation time: 10 min) is relatively thin (estimated at 150 nm). Figs. 9, 10 and 12 show the profiles of the different ions for a conversion coating which is manufactured under the same conditions as the typical layer but for a preparation time of 60 rain (evaluated at 600 nm at least by means of the total ionic sputtering time). Similar profiles are shown by both coatings. Obviously the changes observed in these profiles are clearer for the thicker coating, consequentially comments are made for the latter. The analysis of the results suggests that the conversion coatings are composed of 2 sublayers which are covered by a very thin film. These sublayers are situated in the deep zone. The thin films probably, results from coating degradation by complex mechanisms involving dissolution, oxidation and even deposition reactions. It is a

superficial

/externa.L • nternal , zone ?.zone

Intensity (cts.s-1) 10.10'

~

104

0

j ~-

~

.~Fe+)

3 . 10 -1 2 . 1 0 -1

-,~~ . , f J. . . . . . Fe 3 O + x 3 0

~-':--..::==~'-"~-" "-"~Ze--+-~L _

0

Fe30+), 4 . 1 0 -I~l(Fe*)

I(Fe30+),

4

-4..~14u

2.

1 steel

t

8.104 6.10

film

1 . 10 -1

= ! a I I I ' 0 1200 2400 3 6 0 0 4 8 0 0 6 0 0 0 T i m e (s)

Fig. 12. SIMS intensity of Fe + and Fe3 O + and relative intensity l ( F e 3 0 + ) / l ( F e + ) versus sputtering time for a thick coating.

L. Aries et al. / Analysis of the coating on ferritic steel

152

privileged exchange region with the bath - we detected the presence of numerous secondary ions of metals and alkaline metals which seems to be present at a lower concentration in the coating. The superficial film may also contain compounds or elements which have diffused in from the bath. But from the OH profile it is also plausible that the film contains oxides, hydroxides and adsorbed water molecules. Finally it seems that the thickness of the superficial film is unchanged by the treatment time and remains constant whatever is the total thickness of the coating. Both sublayers may be pro parte composed of iron and chromium oxides. The variation of the profile of the relative intensity l(Fe30+)/I(Fe +) suggests that the external zone contains oxide, hydroxide or oxyhydroxide of Fe -~+ probably associated with Cr 3+, whereas the internal zone contains spinels oxide (Fe304 type) probably corresponding to chromium substituted magnetite. The ionic intensity C 2 which is particularly high within the coating may be due to the presence of carbides or organic compounds. The profiles obtained do not form distinctive steps but curved variations from one zone to the other. This aspect may have different origins. The different strata do not have a perfectly homogeous composition or density and rather there exist gradients. Moreover the roughness of the material may be of some importance, although the size of the analysed zone (about 100 ~m diameter) limits the influence of this phenomenon.

5.2. Electron spectroscopy for chemical analysis (ESCA / XPS) 5.2.1. Analysis of the superficial zone of the conversion coating (fig. 13a) Identification of the principal compounds Oxides and hydroxides The oxygen ls spectrum presents two very large peaks at 530.4 and at 531.8 eV. The first one is representative of oxygen bound to a metal in an oxide [23, 25-29]. The second one may be attributed to oxygen of absorbed molecules such as CO and H 2 0 (532 eV) because of a superficial contamination or to the OH group (531.8 eV) in hydroxides [18,30,31]. The spectrum of iron 2p clearly shows 2 main peaks with energies of 711.2 eV (Fe 2p 3/2) and 724.8 eV (Fe 2p 1/2); satellite peaks at more than 8.5 eV being on the high energy side confirm the existence of Fe 3+ corresponding to F e O O H or Fe203 or chromium substituted corresponding compounds. The presence of hydrated oxyhydroxide cannot be excluded [15-20]. The chromium 2p spectrum shows the characteristic peaks of Cr 3+ at 576.9 and 586.7 eV; these peaks are generally associated with Cr203 [21-24], but solid solutions between chromium (Cr 3+) oxide or hydroxide and iron (Fe 3+) oxide or hydroxide may be present. Sulfates and sulfides The sulphur 2p spectrum shows two peaks of a low intensity at 169.1 and 162.2 eV which are attributed to the species SO42- and S 2 [33-35]. It is not possible to determine the nature of these compounds, nevertheless the ESCA peaks of FeS and Fe203 (Fe 2p 3 / 2 ) are very close and FeS may be present in the film. Identification of minor compounds -

I

Intensity (a.u.)

I

0 is

1

I

Binding energy

d

c

b

a

73O

I

Intensity (a.u)

I 720

-

I 710

I

Intensity (au.)

Fe 2p

I

I

Binding energy I 700 ( e V )

~ d

~ c

a

i

a

[9

I

Binding energy

Cls

I 590

Intensity (e,u.)

Nb 3d

I

d

B

i

I

560

S2p

(eV)

I Bindingl energy 570

Intensity (a.u.)

51)0

I

Cr 2p

I

I

Binding energy

Si2s

154

L. Aries et al. / Analysis of the coating on ferritic steel

The niobium 3d spectrum shows 2 main peaks at 207.2 (Nb 3d 5/2) and at 209.9 eV (Nb 3d 3/2) which are characteristic of Nb 5+ [32]. The silicon 2s spectrum reveals one single peak at 154 eV. This peak can be attributed to SiO 2 [32]. The carbon ls spectrum is very complex involving several peaks (284.8, 286.3 and 288.8 eV). It may be related to the presence of carbon oxide and hydrocarbons in the coating, owing to a contamination effect [33 36].

5.2.2. Analysis of the deep zone The deep zone can be reached after ionic sputtering for 2 to 10 min. (fig. 13b, c). Identification of the main compounds. The oxygen ls spectrum presents a peak at 530.4 eV characteristic of oxygen bound to a metal in an oxide. The iron 2p spectrum presents peak at 709.9 and 723.2 eV which correspond to Fe 2+ going along with satellite structures at 5.8 eV on the high energy side, this spectrum is generally attributed to Fe~O4 (magnetite) [15-20,26]. Since the main peaks of Fe 3+ and those of Fe 2+ are very close to one another and owing to the fact that there is a parasite reaction due to the impact of the ion beam (the Fe 3 + species are reduced to Fe 2+) [37] Fe 3" must coexist with Fe 2 + according to the hypothesis of existence of magnetite probably substituted by Cr 3+. The spectrum of 2p iron, presents peaks at 707.4 and 720.5 eV which correspond to metallic iron probably alloyed with chromium [15,17 20,26,38,39] or to FeS [32]. After a sputtering time of 2 to 20 min the main peaks attributed to Cr 3+ (576.9 and 586.7 eV), coexist with 2 characteristics of the peak of metallic chromium at 574.5 eV (Cr 2p 3/2) and 583.8 eV (Cr 2p 1/2) [40 42]. Chromium sulphides or carbides cannot be excluded. The peaks of Nb 5+ (207.2 and 209.9 eV) decrease in intensity as the coating is penetrated and at 204.1 eV (Nb 3d 5 / 2 ) and 207.6 eV (Nb 3d 3/2) there appear peaks which are attributed to niobium carbide [32]. The intensity of the characteristic Si 2s peak decreases deep in the coating and may be identified as SiO 2. Only the sulphide form remains on the S 2p spectrum with a peak at 162.2 eV. These compounds probably correspond to nickel and iron sulphides.

5.2. 3. Elementary analysis of the coating Table 3 indicates the atomic percentage of the element studied, without carbon

Table 3 Elementary composition of the typical conversion coating after various sputtering times (XPS analysis) ") Sputtering time

Surface 2 min 10 min

Composition (atomic %) S

O

Cr

Fe

Ni

Cu

Nb

2.7 2 0.9

66 44.9 23.1

1.9 5.7 10.5

27.1 43.1 64.2

0.5 1.9 0

0.2 0.2 0

1.5 1.9 1

a) These results do not take carbon into account as its concentration is not significant.

L. Aries et al. / Analysis of the coating on ferritic steel

155

Fig. 14. Dark field image of a typical selective coating showing the diffractant areas corresponding to the diffraction pattern fig. 15.

Fig. 15. Electronic diffraction pattern of an area of a typically treated

specimen.

156

L. Aries et al. / Analysis of the coating on ferritic steel

the value of which is not significant. After 10 min sputtering the composition is still not that of the substrate. Iron is the preponderant element, with a 50% atomic content in the coating, while oxygen and chromium almost complete the composition.

6. Identification of the crystallised compounds Fig. 14 shows an electron micrography of thin foil treated in typical conditions. The diffracting zones are numerous and not very big (about 0.1 #m), they seem to be divided up in a homogeneous manner in the coating. Fig. 15 represents an electron diffraction pattern. This pattern is composed of points and rings; several diagrams of the same type were obtained at different places over the surface and for various angles of the sample. The detailed analysis of this pattern is given in table 4. The points may be the result of the diffraction of crystals that are relatively big, probably corresponding to steel, The rings correspond to micro-crystals of magnetite type.

Table 4 Numerical data corresponding to the electron diffraction pattern taken from a typically treated specimen (fig. 15) ~) Measured values

ldentificated compounds Fe304 b)

Type

D (mm)

dhk ~ (pm) ~)

~d (pm)

ring

15.5 18.5 26.1 32.0 37.0 41.5 48.5

246 206 146 119 103 92 79

12 9 5 4 3 3 2

point

18.48 49.00 55.50

206 78 69

9 2 2

hkl

Steel

dhk I (pm)

hkl

1/1 o

253.1 209.8 148.4 121.2 104.9 93.8 78.2

311 400 440 444 800 840 953 1042

100 28 43 4 5 3 0.4

110 231 141

dh~ I (pm)

hkl

1/1 o

203.0 76.7 67.7

110 321 411

100 5 2

Angle

Measured value ( o )

Theoretical value ( o )

[11011141] [110][231]

59.2 78.5

60.0 79.11

a) D: diameter of a ring or distance between points on the pattern, dhkl: spacing of lattice planes (hkl). Ad: interplanar spacing uncertainty, hkl: Miller indices. I/Io: intensity. The angle values are related to the indexing of points pattern. b~ Probably Cr substituted. ¢) 1 p m = 1 0 12 m.

L. Aries et al. / A n a l y s i s of the coating on ferritic steel

157

7. Discussion-conclusion

The different analyses of the conversion coatings reveal their complex nature. On the one hand we can identify strata which have a difference in the cohesion and in the chemical composition; on the other hand there are numerous chemical compounds present in various crystallization states. Fig. 16 shows a tentative phase representation of the typical coating drawn from all the analytical techniques used. The width of the domain of given phases, at a given depth, is proportional to the ratio of the number of metal atoms present in these phases to the total number (only for the main compounds). This scheme allows the relative importance of the components to be shown at various depths. There are five domains, and it is possible to distinguish 3 zones according to the depth: the superficial film (A), the external (B) and the internal (C) zones which both form the deep zone. - the superficial film: its adhesion to the coating is quite weak. Its thickness may be estimated at 20 nm or thereabout. It is mainly composed of iron (Fe 3+) and chromium (Cr 3÷) oxides or hydroxides or hydrated oxy-hydroxides. Nickel is more abundant than in steel and probably exists in the form of sulphide or sulphate. Many minority products like sulphide, sulphate, water molecules, hydroxides, steel constituents (other metals, niobium compounds), bath components (alkaline metals) are present in this film.

%IV."

I .

D E P T H

li

/

S P U T C T E R I N G T I ME

/

/ /

/

/

/

/

/

/

/

.

.

.

/

/ COATING

J

.i

/

STEEL ATOMIC PROPORTION

Fig. 16. Representation of the composition of the typical selective coating: the scheme gives the atomic proportion versus the depth. Atomic proportion is the ratio between the number of metal atoms in the different compounds to the total number of metal atoms. (A): the superficial thin film, (B): the external zone (deep zone), (C): the internal zone (deep zone), I: domain of Fe 3+ and Cr 3+ oxide and hydroxide, II: domain of - probably Cr 3÷ substituted - magnetite, III: domain of metallic iron and chromium (alloy), IV: domain of metal sulphate(s), V: domain of metal sulphide(s).

158

L. Aries et al. / Analysis of the coating on ferritic steel

- the external part of the deep zone: the p r e d o m i n a n t phase is a crystallised oxide phase of m a g n e t i t e type p r o b a b l y substituted by Cr 3+ and metallic particules of steel. Sulphides (p ro b a b ly N iS and F e S : ) and sulphates are slightly less frequent than in the superficial film. Some m i n o r c o m p o u n d s are also present, i.e. n i o b i u m carbides, and hydrated molecules. T h e thickness of this zone is in the region of 35 nm. - the internal part of the deep zone: it is mainly c o m p o s e d of metallic iron and metallic c h r o m i u m , p r o b a b l y c o r r e s p o n d i n g to the steel. Crystallised oxide phase of m a g n e t i t e type is also found. T h e thickness of this subcoating is in the region of 100 nm. T h e present work has given us a fuller u n d e r s t a n d i n g of the conversion co at i n g used as solar absorber. T h e particle dimensions of each phase h o w e v e r and their distribution t h r o u g h o u t the conversion coating remain to be defined in detail. F u r t h e r studies should be d e v o t e d to m o d e l i n g the co at i n g in order to explain the selectivity (and the a, and ~ values) and to investigate the m e c h a n i s m s of f o r m a t i o n of this conversion coating.

Acknowledgements Th e S E M and T E M analysis have been p e r f o r m e d in the " L a b o r a t o i r e d ' O p t i q u e E l e c t r o n i q u e " ( C N R S Toulouse); Professor B. Jouffrey, Director, M. D a n g and M. Crestou are gratefully acknowledged.

References [1] J. Spitz, Thin Solid Films 45 (1977) 31. [2] R.E. Hahn and B.O. Seraphin, Physics of thin films, Spectrally selective surfaces for photothermal solar energy conversion (Academic Press, New York, 1978) vol. 10. [3] Selective surfaces, Sun II, Proc. Inter. Solar Energy Soc., Silver Jubilee Cong., Atlanta, Georgia (1979) eds. K.W. BOer and B.H. Glenn (Pergamonn Press, New York, 1979) vol. 3, pp. 1885-1938. [4] O.P. Agnihotri and B.K. Gupta, Solar selective surfaces (Wiley, New York, 1981). [5] U.R. Lenel and R.P. Mudd, Solar Energy 32 (1984) 109. [6] T. Karlsson, E. Valkonen and B. Karlsson, Solar Energy 34 (1985) 417. [7] L. Aries and J.P. Traverse, Proc~d~ de fabrication d'un absorbeur s~lectif de capteur solaire et absorbeur s~lectif obtenu, Brevet FR. ANVAR 79 18 414 (1979) US Patent no. 44 05 414. [8] L. Aries and J.P. Traverse, Proc~d~ de fabrication d'un absorbeur s~lectif de capteur solaire et absorbeur s~lectif obtenu, Brevet FR. ANVAR 81 13 815 (1981) US Patent no. 44 44 600. [9] L. Aries, J.P. Bonino, R. Benavente, A. Laaouni and J.P. Traverse, J. Physique 42 (1981) C1-213. [10] L. Aries, .P. Bonino, R. Benavente and J.P. Traverse, Mater. Res. Bull. 18 (1983) 781. [11] L. Aries, F. Berrekhis, R. Benavente, J. Bonino and J.P. Traverse, Mater. Res. Bull. 18 (1983) 1113. [12] L. Aries, Y. Baziard, D. Fraysse and J.P. Traverse, Solar Energy 36 (1986) 6. [13] R. Namdar-lrani, J. Microsc. Spectrosc. Electron. 2 (1977) 293. [14] B. Alexandre, R. Berneron, J.C. Charbonnier, R. Namdar-lrani and L. Nevot, M~m. Sci. Rev. M&allurgie 9 (1981) 483. [15] C.R. Brundle, T.J. Chuang and K. Wandelt, Surf. Sci. 68 (1977) 459.

L. Aries et al. / Analysis of the coating on ferritic steel

[16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42]

159

Masaoki Oku and Kichinosuke Hirokawa, J. Electron Spectrosc. Relat. Phenom. 8 (1976) 475. N.S. Mac Intyre and D.G. Zetaruk, Anal. Chem. 49 (1977) 474. G.C. Allen, M.T. Curtis, A.J. Hooper and P.M. Tucker, J. Chem. Soc. Dalton Trans. (1974) 1525. J.P. Coad and J.G. Cunningham, J. Electron Spectrosc. Relat. Phenom. 3 (1974) 435. J.E. Castle and M.J. Durbin, Carbon 13 (1975) 23. J.E. Holliday and R.P. Frankenthal, J. Electrochem. Soc. 119 (1972) 1190. K. Asami, K. Hashimoto and S. Shimodaira, Corr. Sci. 18 (1978) 151. G.C. Allen, P.M. Tucker and R.K. Wild, J. Chem. Soc., Faraday Trans. I1 74 (1978) 1126. R. Konishi and S. Kato, Jpn. J. Appl. Phys. 14 (1975) 1467. G.B. Smith and H. Ignatiev, Solar Energy Mater. 4 (1981) 119. A.G. Akimov, I.L. Rozenfel'd, L.P. Kazanskii and G.V. Machavariani, Izv. Akad. Nauk. SSSR, Set. Khim 6 (1978) 1239. T. Dickinson, A.F. Povey and P.M.A. Sherwood, J. Chem. Soc., Faraday Trans. 73 (1977) 327. R.L. Chance and S.W. Gaarenstroom, Corrosion 36 (1980) 94. J. Haber, J. Stoch and L. UnDer, J. Electron Spectrosc. Relat. Phenom. 9 (1976) 459. N.W. Roberts, Surf. Sci. 62 (1977) 431. J.K. Gimzewski, B.D. Padalia and S. Affrosman, Surf. Sci. 62 (1977) 386. D. Briggs, Handbook of X-ray and ultraviolet photoelectron spectroscopy (Heyden, London, 1977). P. Marcus, J. Oudar and I. Olefjord, J. Microsc. Spectrosc. Electron. 4 (1979) 63. B.A. Baldwin, Am. Soc. Lubr. Eng. Trans. 19 (1976) 335. R.A. Walton, Coord. Chem. Rev. 31 (1980) 183. G. Nyberg, Surf. Sci 82 (1979) 165. L.I. Yin, S. Ghose and I. Adler, Appl. Spectrosc. 26 (1972) 355. G. Ertl and K. Wandelt, Surf. Sci. 50 (1975) 474. K. Asami and K. Hashimoto, Corr. Sci. 17 (1977) 559. P. Aubrun, G.A. Pennera, C.G. Legras and D. Courteix, J. Microsc. Spectr. Electr. 1 (1976) 43. Y. Okamoto, M. Fujii, T. lmanaka and S. Teranishi, Bull. Chem. Soc. Jpn. 49 (1976) 859. S. Storp and R. Holm, Surf. Sci. 68 (1977) 10.