463

DIFFUSION OF I{ELIUM THROUGH QUARTZ : RELATION

TO TEMPERATURE1
Bv E. F. Bunror,2 E. O.

BR-A.AIEN,3 aNo

J. O. Wrlnerua

Abstract
A qualitative study of thc diffusion of ccrtain

madc lor ordlnary tempcraturcs and prcssures.

gases

through quartz has been

Can. J. Res. Downloaded from www.nrcresearchpress.com by Oregon State University on 12/29/14

For personal use only.

nearly linear relation *'as found to exist between diffusion and pressure
. Aconstant
lor
temperature and a form of thc equation satisfying the cxpeiimental
results suggested. Temperaturc-diffusion curves rvere obtained arrd a c.--

parison made ro previous-worl<. Permeabiliry to helium of single quartz crystalIine plat.s is independent of rhe dirccrion in u'hich rhe platc hIs bien cut, 6uL is
only about onc-half the value of that of fuscd quartz bulbs or platcs.

Introduction
Considerable work has been done on the measurement of the passage of
various gases into or through solids at high temperatures; in particular the
curious action ol quartz towards helium and other gases has received a good
deal of attention.
In 1900 villard (11) found that silica glass was permeable to hydrogen at
high temperatures; a little later Jacquerod and Perrot (2) observed that
helium leaked out of a thermometer. Since then many r,vorkers have obtained
both qualitative and quantitative results for the sorption of gases, both by
glass and metals. Mayer (6) observed the leakage of hydrogen, nitrogen,
and' oxygen through quartz at high temperatures and for three different
pressures. Richardson (7) and also many others (3,8, 12, 13,14) studied the
passage of hydrogen and helium through different kinds of glass. In recent
work by T'sai and Hogness (10) rates of diffusion, between 180o and 955o c.,
were obtained for helium passing through quartz and, from 520' to 9g0o c.,
for neon, also through quartz. In general it has been found that the rate of
diffusion is directly proportional to the pressure and an exponential function
of the temperature (+, 9).
Baxter and Starkweather (1) and Mclennan and wilhelm (5) found that
diffusion of helium through quartz takes place also at ordinary remperarures
and at low temperatures.

Experimental
The work reported in this paper was undertaken in order to investigate
carefully and quantitatively the following:(1) the diffusion of different gases through.quartz at ordinary temperarures;
(2) the variation in the diffusion of helium through quartz with pressure, at
constant (room) temperature; (3) the variation in the diffusion of helium
through quartz with temperature at constant (atmospheric) pressure; (a) the
diffusion of helium through crystalline plates of quartz, and a plate of fused
quartz at normal pressure and temperature.

L Monuscript recaiaed, April 3, 193J.

Contribution from the Physicat Laboratories, (Jniaersity of Toronto, Canad,a.
^ Head'
2
of the Department oJ Phys'ics and' Director of the-MZLennan Laboratory,

(Jniaersity

oJ Toronto.

3.

flol(er.oJ.a Graduate lelowlhip in the Grad.uate School oJ the Unhtersity of Toronto.

Technicol and Research Ass'istan!, i,n the Department oJ Physics, University'oJ Toronto.

CANADIAN JOURNAL OF RESEARCH

Apparatus and Procedure

Although helium gas passes fairly readily through crystalline quartz, fused
quartz n*l Py."* glass, it is not able to pass through ordinary soft glass to
any noticeable extent. consequently the general method adopted in these
experiments was to maintain a pfessure of helir-rm gas inside a quartz bulb
which was sealed into an outer chamber of soft glass'

Can. J. Res. Downloaded from www.nrcresearchpress.com by Oregon State University on 12/29/14

For personal use only.

t<
Ail
I li

/ ll =F
\
\ \l I
).' l\ /
,'rt*o s*s"-z-! \ \\( |

Fig. 1 shows the arrangement

provided for mounting the fused
quartz bulb as used in most of
sasrnbt itr" experiments. T6e quartz container had a total area of 207 sq'
tm', the bulb Part measuring 3'o

cm' in diameter and 0'80 mm' thick'

the neck 1 '80 cm. in diameter and
1'36 mm' thick. The quartz tube
was fitted with a wax-sealed ground
glass joint into a soft-glass container
r,vhich was connected to a Mcleod
and diffusion pump. The
"i;:;!:::11:'" gauge
quartz bulb was kePt filled with
Quartz helium at a given pressure ancl the
sorts/ass soft-glass chamber initially pumped
out to a lou' pressure. Several
investigators have shor'vn that the diffusion of helium through soft glass
is small, even for high temperatures,
so that losses of helium from the low
Frc.1. ApparatususedJor quartzbulb'

i"r"..ur"

system .n,ere negligible.

The thickness of the quartz was obtained by three methods: (a) by

the method of volumes, e-.g., the difference between the volume displaced
on submerging the bulb in a liquid and the volume required to fill the
bulb woulJ give the average thickness when divided by the total area;
(b) bv weight, assuming ttre density ol quattz as 2'2; (c) with a microscope by focusing on the inner and outer surfaces and correcting for
the apparent decrease in thickness due to refraction. The three methods
checked very well.
Fig. 2 gives the arrangement for a plate of crystalline quartz' Two different
plute". *ie obtained fro* a large single crystal, one cut parallel to, the other

axis. Each was 3'5 cm' in diameter and 1'80

'I'he
surfaces u,-efe ground with very fine emery and slightly
mm. thick.
polished with rouge. A small fused quartz plate of the same thickness and
2.5 cm. in diameter, and having received the same surface treatment was

perpendicular to, the optic

also investigated.
Red sealing wax was used for sealing the quartz to the soft glass containers'
The pressure was measured on a Mcleod gauge which could be read to

DII''].USION

O]?

IIELIUM THROI]GII QUARTZ

46'5

Can. J. Res. Downloaded from www.nrcresearchpress.com by Oregon State University on 12/29/14

For personal use only.

0.1 x 10--a cm. of mercury and a thermostat control served to maintain

an oil bath at the required temperature. The temperature of the gas
was measurecl with a thertnocouple inside the quartz bulb'
The low pressure system into which the gas diffused
was first thoroughly evacuated and tested for leaks'
The high pressure side was then pumped out and filled
u,ith helium to the required pressure' after first washing
this high pressure side of the system two or three times
rvith that gas. Readings were continued at a definite
temperature ancl pressure until a constant rate of
increase in pressure r'vas observed, which in some cases
took several days.
The rate of diffusion was obtained by calculating
the number of cubic centimetres of gas diffusing through
per hour per unit surface o{ wall 1 mm' thick, for a
clifference in pressure of one atmosphere' Since the
temperature was not the same over the u'hole low

made' This was easily

removing the bath and bringing the entire
system to room temperature. The two corresponding
readings of the pressure gave the correction factor'

9os

Inlet

Qtartz Plale

oft 9/ass

pressure system corrections rn'ere

clone by

To

/'lcleod 9a7e

Frc. 2.
used.

lor

AqParatus

qtr'artz Plates.

Results
1. The rzrtes of diffusion of air and the two inert gases, neon and argon'
the rate for
\\-ere too small to be measured at ordinary temperatures, r'vhile
hydrogen was just observable'
TABLE I
RBT-ATTON BETWEEN

TIIE DIFFUSION RATE OF HELIU}'I

TEMPERATIRE

Pressure. cm. (b

Rate of diffusion'

.4
23.4

59.0
17'05

77

Xfi6

(tt

GAS AND THE PRESSURE AT CONSTANT

(21'C')

38.6
r0.76

19.2

4.98

5.0

t'+7

2. In Table I are recorded the results shorn'ing the relation between

ffi
Frc.

3.

L_\\Z
I r "r/l

Curtse showittg relcttiot't between'

sion ontl pressure at constunt temperoture.

di,fu-

the diffusion rate of helium gas and

the pressure at constanL temperature
(21" C.) ; the apparatus is that shown
in Fig. 1. ln Column 2 the rate of
ciiffusion is calculated in cc. per hour
per sq. cm. surface of quartz Per
mm. thickness.
The curve in Fig. 3 shorvs graPhically the relation between rate of diffusion and pressure. The equation
r : k tP t krf'rePresents the curve,
where Ar ='= 0'15 and kz : 0'0007'

CANADIAN JOURNAI. OF RESEARCTI

TABLE

II

VARIATION OF RATE OF DIFFUSION OF HELIUM TIIROUGH

QUARTZ WITH CHANGE IN TEMPERATURE

FRor{ +110o to
-185. C..q,r ArnospuERrc

Burton, Braaten and

Temp.,

78
195
LJZ

Can. J. Res. Downloaded from www.nrcresearchpress.com by Oregon State University on 12/29/14

For personal use only.

253.

and Hogness
Log

0.87 x 10-0
Z.JI
5 .02
13

294

23.0

383

121.6

Rate of diffusion

-6.0605

-5.1221
27
- 4.87
-4. 6383

7t3

808
858
923
1043

-3.91.51

900

2,100

Log

-3.523
-3.046

-2.678

3 ,306

-2.

5,100
7 ,200

-2.293

4,200

153

9,400

r228

11,300

-1.

482

L+.>

-2.O27

-1.947

3. The results showing the

variation of the rate of diffusion of helium through quartz
with change in temperature from

+110' to -185" C. at atmo-

spheric pressure are recorded in

Table II and illustrated graph-

.s

300.0 x 10-6

453
583

-5.7100
.2993

.38

273

PRESSURE

ically in Figs. 4 and 5. It $'ill

be apparent from the figures
that a sudden increase in the
rate of diffusion sets in at about
- 30o C. as we proceed upTupe rot re oAbso/ule
wards from the lowest temper{tg 4. Curae showing relot,ion between d,ifusion
atures; the rate then continues
and ternoera.ture d,t conslant pressure,
to increase more rapidly rvith
increasing temperature. In Fig. .) tne temperature rvas plotted against the
logarithm of rate of diffuu

sion. As a means of comparison the results of other

investigations $'ere

plotted

on the same graph w-ith $

the same scale. After the t
sudden increase in perme- $
ability at -30o C., it then t

continued fo i".r*;;;

rapidly for higher temperatures according to an

exponential law up to 80'
or 90o C. Williams and
Ferguson found an
exponential law for high
temperatures while arr ex-

o-Wiuiams LFe/qusaa - IZ Ho?.ess

o^-Ttai
-Bu,ton, 6tuatea L Wilhe/n

Frc. 5.

Sanne as F,ig. 4 showing

lemperature.

log (d,ifusion)

and,

trapolation of the authors'results fits in well n-ith the work of Tsai and Hogness.

DIFFL:SION OF IIELIUM THROT]GH

QUARTZ

467

4. Diffusion through ground plates ol quartz'

there is a difference between fused and crystalline

as follows
quaftz,measurements were made on three ground plates of quartz,
mm' in
1'85
and
diameter
in
(see Fig.2): (1) A crystalline plate,3'5 cm'
plate
3'5
(2)
A
crystalline
axis;
thickness, cut perpendicular to the optic
axis;
optic
the
to
parallel
cut
cm. in diameter and 1.79 mm. in thickness,
(3) A Iused quartzplate ground similarly to the others' 2'5 cm' in diameter

Can. J. Res. Downloaded from www.nrcresearchpress.com by Oregon State University on 12/29/14

For personal use only.

In order to fin<l whether

and 1 .75 mrn. in thickness.

cc' per
The resulting rates of dilTusion were as follows: (1) 11'8 X 10-u
per
cm'
sq'
per
hour
hour per sq. cm. per mm. thickness; (2) 12' 1 X t0-u cc'
thickness'
per
mm'
p", -'*. thickness; (3) 28 3 X 10-(] cc' per hour per sq' cm'
It is interesting to note that the fusecl quartz ground plate. gave only a
slightly greater rate than the fused quartz bulb with smooth transparent
of the
surfaces; also, that the rates for the crystalline plates are independent
for
the
value
direction in rvhich the plate is cut and is only about one-half
fused quartz.
which will be
These results give rise to interesting theoretical speculations
performed'
been
have
experiments
reserved for a future paper, r,vhen furiher
References
G. P. and Srlnr<rvBarunR,J{' $r' Science' 73:618' ^1931'
j:1. BexrBn,
pu*nor, p' L' com.pt' rend' l3e: 78e' 1e9s'
i.i;;;;;.;,j: A.;.a
12))'
Fi. ;;J siii' H.-c.-1. bpiii"t Soc' Am' 6: 734-738'
l: i;;I';i,
Geo. Routledge
Tni,i,pi"",ir

;;,;;;;a".;ois'by

;. d;B;;N.'i.- ii;.-

London. 1932'
5. McLrnNeN, J' c. u"iiiiittulln., |.O .llo-'' Rov'
o. ltou.*, tr' ( . Phvs. Rev.6:283-291' 1q15'
Ltd..

"olids.

Soc' Can'

and Sons,

III 26:119-122'

7932'

7. RrcnenosoN, O. \\i. u"Jfiti"i"otox' R C Phil' NIag' 22-:704-706.-1911"

A Ir7:662-679,
8. Sro.rcrn, E. W. R. ;;'i";;;;;"F. M. qr ero.. n.y.-so.. Lond.
9. T;;;.. d";aday soc. is, tz9-++7 ' 1932'
36:2595-2600' 1932'

10. T'ser,

Lru-Snexo,."ifr'ooi"it, T.-R,-J^. Phvs' Chem'

re00'

ii: Vt;;"",-P.

Conipt. rend. 130: 1752-1753' .

12. \\irl,rr.A'Ms, G. A. and
l' -P' J
13. \Urrr-r"tus, G. A. and'i;;.;;;;:j^.^g -.i'^

it'iiiJi'i"'

Chem' Soc' 44:2t60-2167' 1922'

S''
{q Chem' Soc' 46:635-63e' re24'

i,i. iviitt-"t", A. e""' Phvsik,46: 1095-1129'

1915'

1928.