Hoffman D.M., Singh B., Thomas J.H. (Eds). Handbook of Vacuum Science and Technology

Подождите немного. Документ загружается.

532

Chapter 4.8: Aluminum-Based Vacuum Systems

Table 5—Continued

DESIG-

NATION

6003®

6005

6053

6061

6063

6066

6070

6101®

6105

6151

6162

6201

6253®

6262

6351

6463

6951

7005

7008®

7049

7050

7072®

7075

7175

7178

7475

8017

8030

8176

8177

SILICON

0.35-1.0

0.6-0.9

<S)

0.40-0.8

0.20-0.6

0.9-1.8

1.0-1.7

0.30-0.7

0.6-1.0

0.6-1.2

0.40-0.8

0.50-0.9

<S>

0.40-0.8

0.7-1.3

0.20-0.6

0.20-0.50

0.35

0.10

0.25

0.12

0.7 Si

0.40

0.15

0.40

0.10

0.03-0.15

0.10

IRON

0.35

0.7

0.35

0.50

0.35

1.0

0.50

0.7

0.50

0.15

0.8

0.40

0.10

0.35

0.15

+ Fe

0.50

0.20

0.50

0.12

0.55-0.8

0.30-0.8

0.40-1.0

0.25-0.45

COPPER

0.10

0.15-0.40

0.10

0.7-1.2

0.15-0.40

0.10

0.35

0.20

0.10

0.15-0.40

0.10

0.20

0.15-0.40

0.10

0.05

1.2-1.9

2.0-2.6

0.10

1.2-2.0

1.6-2.4

1.2-1.9

0.10-0.20

0.15-0.30

o!64

MAN-

GANESE

0.10

0.15

0.10

0.6-1.1

0.40-1.0

0.03

0.10

0.20

0.10

0.03

0.15

0.40-0.8

0.05

0.10

0.20-0.7

0.05

0.20

0.10

0.30

0.10

0.30

0.06

MAG-

NESIUM

0.8-1.5

0.40-0.6

1.1-1.4

0.8-1.2

0.45-0.9

0.8-1.4

0.50-1.2

0.35-0.8

0.45-0.8

0.7-1.1

0.6-0.9

1.0-1.5

0.8-1.2

0.40-0.8

0.45-0.9

0.40-0.8

1.0-1.8

0.7-1.4

2.0-2.9

1.9-2.6

0.10

2.1-2.9

2.4-3.1

1.9-2.6

0.01-0.05

0.05

0.04-0.12

CHROM-

IUM

oli

0.10

0.15-0.35

0.04-0.35

0.10

0.40

0.10

0.03

0.10

0.15-0.35

0.10

0.03

0.04-0.35

0.04-0.14

0.06-0.20

0.12-0.25

0.10-0.22

0.04

0.18-0.28

0.18-0.25

NICKEL

ZINC

olo

0.10

0.25

0.10

0.25

0.10

0.25

0.10

1.6-2.4

0.25

0.20

0.05

0.20

4.0-5.0

4.5-5.5

7.2-8.2

5.7-6.7

0.8-1.3

5.1-6.1

6.3-7.3

5.2-6.2

0.05

0.10

0.05

TITAN-

IUM

oTo

0.10

0.15

0.10

0.20

0.15

0.10

0.15

0.10

0.15

0.20

0.01-0.06

0.05

0.10

0.06

o!20

0.10

0.20

0.06

OTHERS®

Each®

0.05

005

0.03®

0.05

0.03®

0.05

0.05®

0.05

0.05®

0.05

0.05®

0.05

0.03®

0.05®

0.03®

Total®

0.15

0.10

0.15

0.10

0.15

0.10

0.15

0.10

0.15

0.10

ALUMI-

NUM

Mtn.®

Remainder

Note:

Listed herein are designations and chemical composition limits for

some wrought unalloyed aluminum and for wrought aluminum alloys

reg-

istered with

The

Aluminum Association. This list does

not

include

all

alloys registered with The Aluminum Association. A complete list

reg-

istered designations

contained

the "Registration Record

interna-

tional Alloy Designations and Chemical Composition Limits

for

Wrought

Aluminum and Wrought Aluminum Alloys." These lists are maintained

t>y

the Technical Committee on Product Standards

the Aluminum Asso-

ciation.

® Composition in percent

t>y

weight maximum unless shown as

range

minimum.

® Except

for

"aluminum"

and

"others," analysis normally

made

for

elements for

which

specific limits are

shown.

For purposes of determining

conformance

these limits,

observed value

or a

calculated value

obtained from analysis

rounded

off

to the nearest unit in the last right-

hand place of figures used

expressing the specified limit,

accordance

with ASTM Recommended Practice E

29.

® The

sum of

those "other" metallic elements 0.010 percent

each,

expressed

the second decimal before determining the sum.

® The aluminum content for unalloyed aluminum not made by

refining

process is the difference between 100.00 percent and the sum of all other

metallic elements present

amounts

0.010 percent

more each,

expressed

the second decimal before determining the sum.

® Also contains

0.40-0.7

percent each

lead and bismuth.

® Electric conductor. Formerly designated EC.

® Cladding alloy. See Table

6.1.

®Fo«i.

® Vanadium 0.05 percent maximum.

® Also contains

0.20-0.6

percent each

lead and bismuth.

® Brazing alloy.

® Bus conductor.

® Vanadium plus titanium 0.02 percent maximum; boron 0.05 percent

maximum; gallium 0.03 percent maximum.

® Zirconium 0.08-0.20

® Silicon 45 to 65 percent

actual magnesium content.

® Beryllium

0.0008

maxinfHim for welding electrode and welding rod only.

® Boron 0.06 percent maximum.

(3) Vanadium 0.05-0.15; zirconium 0.10-0.25.

® Gallium 0.03 percent maximum; vanadium 0.05 percent maximum.

® In addition

those alloys referencing footnote

®. a

0.0008

weight

percent maximum beryllium is applicable to any alloy to be used as

weld-

ing electrode

welding rod.

® Zirconium 0.08-0.15.

® Includes listed elements

for

which no specific limit

shown.

® Boron 0.04 percent maximum; lithium

0.003

percent maximum.

® Boron 0.001-0.04.

® Gallium 0.03 percent maximum.

® Boron 0.04 percent maximum.

Source:

Aluminum Association,

A/z/mmMm

Standards and Data, Tenth Edition, 900 19th St., N.W.,

Washington, D.C., 1990, p. 97, Table 6.2.

4.8.4 Mechanical Considerations

533

Rg.9.

(Thickness)

Flat plate thickness determination.

P (Atmospheric Pressure)

i 1 i

K-

_ ^

—

(Radius or small

side of rectangle)

—>

_^ (Deflection)

Must limit deflection to an acceptable level.

FLAT PLATES

The thickness of flat plates is normally determined by the maximum deflection

permissible (see Figure 9) and can be approximated as [48]:

where

kPr/

1/3

(2)

^p = thickness of plate

k =

0.696

if undamped, = 0.171 if clamped

P = atmospheric pressure

= radius of circular plate or small side of rectangular plate

E = modulus of elasticity—Al,

7,470

Kgf/mm^;

SS,

19,700 Kgf/mm2

d = acceptable deflection

CYLINDERS

Cylinder walls can usually be much thinner than flat plates. The principal failure

mode for cylinders pressurized on the outside is buckling due to instability of

large unsupported surfaces. (Thickness of cylinders with pressure on the inside

need only keep stresses below tensile strength—see peak stress formulas below.)

The theoretical thickness necessary to prevent buckling in cylinders longer than a

critical length, l^ = 1.11 d^{djt^)^-^ (see Figure 10) is [49]

tr = d.

>(1 -m^)

11/3

(3)

534

Fig.

10.

Chapter 4.8: Aluminum-Based Vacuum Systems

Cylinder thickness determination.

(Pressure)

h<-

Must be sufficient to prevent instability and buckling in large unsupported surfaces (when ex-

ternally pressurized).

where

= thickness of cylinder wall

dc = outside diameter of cylinder

P = pressure (typically 4 atm for safety)

m = Poisson's ratio (approximately 0.3 for both Al and SS)

E = stress/strain at actual operating point (elastic modulus)

/ = length of cylinder or distance between supporting ribs

For cylinders with a length, /, shorter than /c, thickness can be approximated us-

ing the following distillation of graphical and analytical relationships offered by

Brownell (in any consistent unit system) [50]:

0.697

0.402

0.585

/0.414

/O-^

(4)

RELATIVE THICKNESS OF ALUMINUM AND STAINLESS STEEL

Inspection of these formulas demonstrates that the minimum thickness (f^s) ^f

aluminum relative to stainless steel for the same application is related to their

elastic moduli, E, as follows:

4.8.4 Mechanical Considerations

535

r AS (plate) =

^As(Iong cylinder) =

f AS (short cylinder) =

^ss

1/3

"19,400^

7,470

1.37

1/3

= 1.37

^ss

^Al

0.402

= 1.47

(5)

(6)

(7)

Thus,

aluminum chambers must be at most 50% thicker than stainless chambers.

Often other considerations, such as structural stresses or manufacturing con-

venience, may require greater thickness than is necessary to merely withstand

vacuum/atmospheric pressure differential. In such cases, aluminum and stainless

steel may be of comparable thickness.

PEAK STRESSES

AND

STRAINS

Peak stresses should be calculated to ensure that material strength is not ex-

ceeded. These can be approximated as follows:

Prp2

Plate (clamped) stress = 0.75 —j-

Plate (undamped) stress [51] = 1.24

Prr.

Cylinder circumferential stress =

2^7

Cylinder radial stress = P

Cylinder axial (force of end plates) stress [52] =

"4^

(8)

(9)

(10)

(11)

(12)

To ensure that deformation is within acceptable limits, strains should be esti-

mated by dividing these stresses by elastic modulus, E.

WEIGHT AND THERMAL EXPANSION

Aluminum has about one-third the weight density of stainless steel: 2.70 versus

8.03 g/cm-^ respectively

[53].

This makes aluminum appendages much lighter and

easier to support than those made of stainless steel. However, allowances must be

made for its 48% greater thermal expansion: 23.4 [54] versus 15.8 [55] micro-

strain per degrees Celsius respectively.

536 Chapter 4.8: Aluminum-Based Vacuum Systems

4.8.5

THERMAL CONDUCTIVITY AND EMISSIVITY

The thermal characteristics of aluminum are distinctly different from those of

stainless steel and offer significant advantages where hot spots are a problem or

where a chamber must be rapidly heated or cooled.

4.8.5.1 Thermal Gradients

Heat is often concentrated on a chamber wall by sources such as conduction from

a hot appendage or by radiation from a nearby hot filament. The excessive out-

gassing and hazards caused by high temperatures at these points are greatly di-

minished by using aluminum, because it has a thermal conductivity 10 times

higher than that of stainless steel. As an illustration, the temperature rise caused

by a localized heat source on a large flat plate due only to heat conduction (Fig-

ure 11) is [56]

'''' ^ Iwkt

where

T^i^^

= maximum steady-state temperature rise

q = Incident heat

^2 = diameter of outer perimeter at room temperature

di = diameter of heat source (hot spot)

k = thermal conductivity:

ICAI

= 206, kss = 16.2 W/m-°C

t = thickness of surface

For purposes of comparison using the above formula, lOW of heat from a 6mm

(0.25") diameter feedthrough on a

m diameter by 6 mm thick plate will produce

a 6.6°C rise in aluminum and an 83.8°C rise in stainless steel. The temperature

rise is proportional to power and can become substantial for even moderate heat

loads.

4.8.5.2 Absorption of Radiated Heat

Aluminum is highly reflective of incident radiated heat. It has an emissivity in the

range of 0.04-0.07 compared to 0.3-0.7 for stainless steel [57]. In a closed sys-

tem, all reflected heat is ultimately absorbed but in aluminum chambers tends to

be uniformly distributed and not concentrated on the point of initial incidence.

4.8.5 Thermal Conductivity and Emissivity

537

Fig.

II.

Temperature rise caused by incident heat.

Hot Fcedthrough

Trise

(A Temp)

% t (Thickness)

-d,-

(Diameter)

However, heat can concentrate on nonaluminum surfaces with high emissivity if

these have substantial exposed cross sections.

4.8.5.3 Thermal Time Constant

Because of its high thermal conductivity, an aluminum chamber can be uniformly

heated and cooled in a fraction of the time required for a stainless steel chamber.

The thermal time constant of a cylindrical system wrapped with tubing carrying

heating/cooling fluid (Figure 12) can be estimated as follows:

(14)

where

TC = thermal time constant

d = distance between heating/cooling tubes

Cp = specific heat: Cp^i = 870, Cpss = 460 W-sec/kg-°C

D = density: D^i = 2.70, Dss = 8.03 g/cm-^

k = thermal conductivity: kAi = 206, kss = 16.2 W/m-°C

538

Fig. 12.

Chapter

4.8:

Aluminum-Based Vacuum Systems

Thermal time constant.

-d-

Heating/Cooling

Ruid

High thermal conductivity makes rapid heating and cooling of aluminum chambers feasible.

This assumes constant chamber thickness and approximates the distributed nature

of its heat capacity and conductance by a lumped mass at the end of a thermal re-

sistance corresponding to half the maximum distance from the heating fluid. It is

intended to offer a feasibility perspective, not an exact prediction. For purposes of

comparison, if heating/cooling tubes are spaced 20cm (8 inches) apart, the ther-

mal time constants for aluminum and stainless steel are 1 minute and 19 minutes

respectively.

This short thermal time constant of aluminum systems enables rapid partial

baking of commercial systems, especially since, according to Figure 1, aluminum

will start to effectively degas at temperatures corresponding to one time constant

(102°C).

4.8.6

CORROSION

Both aluminum and stainless steel corrode in active gases and have limited life.

Often the practical differences are minimal so that material selection is based on

other considerations. For example, many semiconductor processes interact with

chamber walls to produce particles or vapors. In such cases, aluminum is much

less of a contamination problem than the heavy metals in stainless steel.

4.8.6 Corrosion 539

4.8.6.1 Suggested Approach to Predicting

Corrosion Performance

Corrosion of metals occurs by many different chemical mechanisms, making the

process so complex and difficult to characterize that accurate generalizations are

nearly impossible. Nevertheless, a designer can use published corrosion data and

a basic understanding of the mechanisms to make

reasonable

guess.

Performance

must then be confirmed by testing under actual operating conditions. Published

corrosion rates of aluminum and stainless steel for the most chemically active

gases are given in Table 6 [58-79].

4.8.6.2 Principles for Extrapolating Published Data

The effects of other corrosive gas compounds is generally no worse than that of

their halogen component alone. The principal mechanism of gas corrosion in-

volves the scale (oxide layer) functioning as a semiconductor. Metal ions and

electrons diffuse through it to the surface where they react with adsorbed corro-

sive gas. The scale grows at the outer surface, not at its interface with the base

metal. This diffusion process appears to be the rate-limiting step in most cases. It

is an activated process that increases with temperature but decreases with scale

thickness [80].

The presence of moisture introduces electrochemical and acid/base mecha-

nisms that usually increase the net corrosion rate. However, this is not always the

case;

for example, moisture greatly reduces chlorine corrosion of aluminum at

high temperatures as shown in Table 6. In general, the transport barrier character-

istics and high electrical resistivity of aluminum scales tend to inhibit the electro-

chemical effects of moisture. However, these scales are attacked if pH falls out-

side the range of about 4 to 8.5 [81]. It has been reported that corrosion begins

when an HX/moisture mixture approaches its dewpoint. For example, HBr at

1 atm with 300 ppm moisture has a room-temperature dew point and causes cor-

rosion; when moisture partial pressure falls below 100 ppm corrosion stops. Dew

point temperature, however, is difficult to predict, because it appears to be posi-

tively correlated to moisture partial pressure and total pressure. Published data for

making such estimates are very limited [82].

4.8.6.3 Heuristic Predictor of Corrosion Resistance

Scales that provide substantial corrosion protection usually have a volume per

mole of metal in the range of one to two times that of the base metal; those out-

540

Chapter 4.8: Aluminum-Based Vacuum Systems

Table 6

Published Corrosion Rate Data Comparing Aluminum and Stainless Steel

Corrosive Gas

Aluminum Corrosion Rate

Stainless Steel Corrosion Rate

Atmosphere For most alloys: <0.1 mm/yr de-

creasing to constant

0.003

mm/yr

after 6 mo-2 yrs in the most severe

sea coast locations [58] (2000 and

7000 series corrode at 2-3 times

this rate [59])

Extremely low [60]

Fluorine

Hydrogen fluoride

For 1100 alloy:

<0.05 mm/yr to 215°C, dry

<0.5 mm/yr to 240^ dry [66]

A durable protective scale forms

[67]

Moisture greatly accelerates

corrosion [68]

2024 and 5224: 0.01 mm/yr

at 547°C

(200 X higher for 1000 scries) [69]

Al-Si Alloys: lower rate [70]

<0.5 mm/yr to 49°C, dry [76]

<4.9 mm/yr to 500°C, dry

< 14.6 mm/yr to 600°C, dry [77]

<0.05 mm/yr at 210°C, dry

<0.5 mm/yr at 249°C, dry [71]

Moisture greatly accelerates corro-

sion [72]

For 316: temperatures slightly

higher [73]

<6.b5 mm/yr to 49°C, dry

<0.5 mm/yr to 204°C, dry [78]

13.4 mm/yr at 600°C, dry [79]

side this range usually do not. This can serve as a useful design tool even though

the reasons for it are not well understood [83].

Thus,

for corrosion protection,

1 <

nW.^ G.

(15)

where

^s,m = molecular weight of the scale and metal respectively

Gs,m = specific gravity of the scale and metal respectively

n = number of metal atoms in the scale molecule

4.8.7 Welding Aluminum for Vacuum Applications 541

This ratio for aluminum oxide (AI2O3) is

(101.94 X 2.70)/(2 X 26.97 X 4.00) = 1.3

which is within the range where corrosion protection is expected, whereas the

same ratio for iron oxide (Fe203) is 2.2 [84], which is outside this range. Iron ox-

ides (rusts) are indeed less protective than aluminum oxides. However, it should

be noted that these occur mainly on carbon steel; the nickel/chromium/iron scales

on stainless steels are highly protective. In general, even though aluminum is very

active chemically, it rapidly forms scales that provide effective passivation against

further corrosion.

One notable exception to the preceding rule is the scale formed by chlorine

(AICI3), which has a ratio of 1.8 but provides limited protection because its high

vapor pressure and sublimation temperature of only 178°C causes it to evaporate

away [85]. In contrast, AIF3 formed by fluorine is known to be highly protective

[86].

It melts at 1197°K and has a vapor pressure of only 0.075 torr at 1145°C

[87].

4.8.6.4 Protective Coatings

Protective coatings of electroless nickel and hard-anodize (thick y-Al203) pro-

vide very effective protection against corrosive gas at moderate temperatures, al-

though the latter tends to increase outgassing by a factor of 40. In a plasma-

activated chlorine environment, these are somewhat less effective (particularly

the nickel). Proprietary, AI2O3 coatings with a high percentage of a-phase have

been reported that virtually eliminate corrosion as a failure mode in turbomolecu-

lar pumps on reactive plasma etch systems [88,89]. A special process is required

to create these, because a-Al203 normally forms above 800°C, which is above

the melting point of aluminum.

4.8.7

WELDING ALUMINUM FOR VACUUM APPUCATIONS

Techniques for welding aluminum are well established and documented

[90,91,

92].

These are somewhat different from those for stainless steel but not much

more difficult. The main objective in fabricating vacuum chambers is to produce

welds free of porosity and cracks that can cause leak paths or virtual leaks. Con-

siderations for structural applications such as strength, color, and corrosion resis-

tance are of secondary importance. The welding of aluminum need not be consid-

ered an art, because mechanisms affecting weld quality follow well-understood

principles.