A New Method for The Determination of Micropore Size Distributions

KK Unger et at. (Editors), Characterization of Porous Solids © 1988 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

89

A NEW METHOD FOR THE DETERMINATION OF MICROPORE SIZE DISTRIBUTIONS

P.J.M. CARROTT, R.A. ROBERTS & K.S.W. SING Department of Chemistry, BruneI University, Uxbridge, Middlesex, UB8 3PH, UK. ABSTRACT A new method of micropore size analysis has been applied to adsorption isotherms of nitrogen (at 77K), propane (at 196K), isobutane (at 261K) and neopentane (at 273K) determined on five well characterised microporous carbons. The method is based on the use of the as method for the characterisation of primary and secondary micropore filling and the determination of the effective micropore volume. The approach, which is still at an early stage of development, appears to offer a number of advantages over other current procedures for the computation of the micropore size distribution. INTRODUCTION The work reported here was carried out as part of a long term study of the

adsorptive

properties

of

a

number

of

well

characterised microporous

carbons. The aim of the work was to extend the application of the as method [1,2] for the analysis of adsorption isotherms of organic vapours and thereby obtain a semi-quantitative estimate of the micropore size distribution. The new method is based on the concept of two stages of micropore filling [3,4,51 (a)

a

primary process occurring wi thin pores of molecular

plpo < 0.01, and (b) ca.

0.2).

dimensions at

a secondary, or cooperative, process at higher plpo

The validity of applying the

as method

to

(to

isotherms of organic

vapours has been confirmed by a careful analysis of data obtained on a range of non-porous carbons and silicas [61.

EXPERIMENTAL Five microporous carbons were selected as representative adsorbents having different pore structures.

Carbosieve S is a polymer based molecular

90 sieve carbon manufactured by Supelco and supplied by Bioscan, Canvey Island, UK. AX21 is capacity

a

petroleum-pi tch-based active carbon of very high adsorption

manufactured

and

supplied

by

the

Anderson

Development

Company,

Michigan, U.S.A. The three samples of charcoal cloth (JF005, JF144 and JF518) were

specially

prepared

from

viscose

rayon

cloth

by

carbonisation

in

N2

followed by activation in CO2 at 850 0C [7]. JF005 is a low burn-off cloth of comparatively low pore volume whilst JF144 and JF518 were prepared at higher burn-off and under conditions designed to produce pore widening and higher BET areas. The nitrogen adsorption-desorption isotherms were determined at 77K using a Carlo-Erba Sorptomatic 1800 which had been calibrated against a manual volumetric apparatus. A Datametrics type 561-91 Barocel transducer gauge was employed

for

the measurement of pressures below 40 mmHg. The

propane (at 196K), isobutane (at 261K) and neopentane 273K)

were measured

with a

CI

isotherms of

(2,2-dimethylpropane at

Electronics Robal microbalance and a

Texas

Instruments Bourdon gauge. Each adsorbent sample was outgassed at 250 0C for 16 hours to a residual pressure < 10- 5 mmHg.

RESULTS & DISCUSSION The Figures

1-5.

adsorption To

isotherms

facilitate

and

corresponding

comparison,

the

as

plots

of

vapour

amount

are

shown

in

adsorbed

is

expressed as the equivalent volume of liquid adsorptive. For this purpose the values of normal liquid density (at the appropriate temperature) of nitrogen, propane, 0.615

g

isobutane

and

neopentane

cm- 3 respectively.

It

were

taken

as

0.808,

is of course appreciated

0.752, that

0.596

and

it cannot be

assumed that the adsorbate actually has liquid-like properties [8]. All

of

the

isotherms

exhibit

some

Type

I

character,

which

is

consistent with the microporous nature of all the carbons, but it is evident that there are some significant differences in their location and shape (e.g.

91

---~ 04

-

"E u

-,

0.2

o. NITROGEN <><>PROPANE o e I SOBUTANE 6"NEOPENTANE

ro

c:

0·2

04

P/p o

0·6

08

05

1·0

20

1-5

o£s

FIGURE 1 Adsorption isotherms and corresponding as plots for Carbosieve. Open points - adsorption. Closed points - desorption.

0·4r--------------,.---------------,

-7' -~¢ --=-.-___ --

0·3

'''' .~

Ci

-~--

0·2

"I:u "-

•

-'

"0

s

•

o.NITROG EN

/'

~PROPANE

e

c:

,-

oerSOBUTANE 6" NEOPENTANE

0·1

0·2

04

P/p

o

0·6

0·8

0·5

1·0

I1S

1·5

FIGURE 2 Adsorption isotherms and corresponding as plots for JF005. Open points - adsorption. Closed points - desorption.

2·0

92

'"" -c

0-4

::J

;;;

~

u

Propane

o e I sobutane

<,

c

/

o. Nitrogen

S!

t1, Neopentane

d

02

02

0-4

0-6

plpo

0-8

05

10

10 o(s

15

20

FIGURE 3 Adsorption isotherms and corresponding as plots for JF144. Open points - adsorption. Closed points - desorption.

1-5.,.....------------y---===-----------,

----_~z~

'1-0


'''''

-5 '" _!:!,

~

.

c

o.Nitrogen .,..Propane

o e I soburene b., Neopentane

0-5

I

0-2

0-4

pi po

0-6

0-8

1-0

/

J!>

0-5

1-0 ols

1-5

FIGURE 4 Adsorption isotherms and corresponding as plots for JF518. Open points - adsorption. Closed points - desorption.

2-0

93

zo-r-----------,---------------,

-- ----~ 7~

15

Lf:.~=~~----

~

'",

"'0 10 5

o. Nitrogen

::

~

Propane oe Isobutane r:" Neopentane

-e-e-

E u

.....

'"

c:

/

05

0-2

0·4

pi po

0·8

06

I

20

15

0·5

10

FIGURE 5 Adsorption isotherms and corresponding as plots for AX21. Open points - adsorption. Closed points - desorption.

scale, rectangularity and reversibility). There is clear evidence of molecular sieve action in the case of Carbosieve and JF005, whereas deviations from the Gurvitsch rule are not so pronounced with the other adsorbents. However,

in

the multilayer region the nitrogen isotherms all lie at a significantly higher level of uptake than the corresponding alkane isotherms. These results suggest that the molecular packing of nitrogen in micropores does not conform to that in

the

liquid

state.

The

low pressure

hysteresis

given by

isobutane

and

neopentane in Figures 1 & 2 appears to be associated with the slow penetration of these bulky molecules into many of the narrow pores present in Carbosieve and JFOOS. The '\ plots in Figures 1-5 have been constructed from the isotherms in the manner described previously [2]. The amount of gas adsorbed is plotted against as' non-porous

the

reduced

carbons

being

standard adsorption on non-porous carbon, used

to

determine

the

standard

data

several for

each

94 adsorptive [6,9]. The thus

s plots are in general linear in the multilayer region;

back extrapolation to

micropore volume, filling

region,

Vo'

the

o

as

available s

plots

to

fall

provides an assessment of the each adsorptive. into

two

Over the

groups

pronounced distortion of the isotherm at low p/po

(a)

initial pore

those

(in Figures

those exhibiting a more gradual approach to the plateau

effective

revealing

1-3)

(Figures 4

and &

5).

(b) As

suggested previously [8], we attribute the first type of behaviour to primary micropore filling of pores of molecular dimensions and the second type to the secondary filling of the wider micropores. If we make certain assumptions concerning the range of pore size which corresponds to each stage of micropore filling for a given adsorptive, we can begin

to

derive

the

micropore

isotherm data obtained so far.

size distribution

from

the complete set of

In slit-shaped pores the limiting width for

primary micropore filling appears to be little more than 20 i.e. two molecular diameters

[10,11].

The

upper

limit

for

secondary micropore

filling

is

at

present uncertain but is likely to be in the region of 50 [11]. On the basis of these assumptions Table 1 gives the limiting pore widths corresponding to primary and secondary micropore filling by nitrogen,

propane,

isobutane and

neopentane. In certain cases it appears to be possible to estimate the primary micropore

volume

unfortunately,

from

the

the DR plots

intercept

of

the

appropriate

DR

plot

for the adsorption of the organic molecules on

TABLE 1 Primary and secondary micropore filling of slit shaped pores.

Adsorptive

Molecular Diameter O/nm

Nitrogen Propane Isobutane Neopentane

0.36* 0.43 0.50 0.62

* minimum dimension 0.30nm.

[6].

Range of pore width, d/nm Primary Secondary Micropore Micropore Filling Filling 0.30-0.72 0.43-0.86 0.50-1.00 0.62-1.24

0.72-1.8 0.86-2.2 1.00-2.5 1.24-3.1

95 JF518 and Ax21 are curved at both high and low pressures and it is therefore not possible to obtain a does

indicate,

however,

reliable value for the intercept. that the

estimate

of

the

provided

by

microcalorimetric

adsorption

[11]

uptake

at

primary micropore volume.

which

have

generally

Additional

measurements

indicated

0. 0 1po

that

of the

The DR analysis gives

a

for

this

is

enthalpies

of

support

differential primary

stage

is

good

virtually

complete at plpo = 0.01. We have therefore taken the uptake at this point on the isotherm to represent the limiting uptake in pores of d = 20. The secondary micropore volume is given by the difference between the total micropore volume,

obtained

from

the

intercept of

the

appropriate

as

plot, and the corresponding primary micropore volume. Note that this does not involve any assumption

regarding the

value of plpo at which the secondary

stage is complete. Values of the total micropore volume and the uptake at 0. 0 1po for each of the isotherms in Figures 1-5 are given in Table 2. pore size distributions,

Using these data the

presented in histogram form in Figures 6-10,

have

been constructed. The diagrams are made up of eight separate entries i.e. two for each adsorptive. The pore size distribution is displayed in the form of vs. d, where 6V o represents that part of the micropore volume filled by

6Vo/~

a particular adsorptive at either the primary or secondary stage. In the case of

Carbosieve

and

JF005

no

evidence

of

secondary

micropore

filling

was

observed with any of the adsorptives and the histograms have therefore been truncated at 0.72nm.

TABLE 2 Total micropore volumes, Vo' and uptakes at 0. 0 1po, Vp. All values in cm3g- 1•

Carbon Carbosieve JF005 JF144 JF518 AX21

Nitrogen Vo Vp 0.43 0.33 0.55 0.98 1. 52

0.43 0.33 0.44 0.43 0.78

Propane Vo Vp 0.36 0.28 0.51 0.90 1. 29

0.36 0.28 0.45 0.43 0.77

Isobutane Vo Vp 0.33 0.25 0.49 0.90 1. 25

0.33 0.25 0.41 0.39 0.71

Neopentane Vo Vp 0.26 0.22 0.49 0.89 1. 22

0.26 0.22 0.42 0.46 0.80

96

1·0

6Vo 6d

r--I

I I

0·6

I

I

I I I

r'-'

I I I

I'

I

I

••

~

'i

,.

I.

0·2

I I

I! I

j

I

I

Ii

I.

0·5

1·0

djnm

1-5

FIGURE 6 Effective pore size distribution for Carbosieve. Solid lines - primary stage. Broken lines - secondary stage.

1-0-

tNo

6d

0·6

r----

I I

I ,•.•• I

I I

I I

I

I

I

0·2

'i I I

,

I'

I

I

'i

I·

0·5

1·0

1·5

d/nm FIGURE 7 Effective pore size distribution for JFOOS. Solid lines - primary stage. Broken lines - secondary stage.

97

lO-

I

I I

I

tNo

I

~._

6d

. . _l_.~

I i i

I I

0·6

I

.

:····t .. r: ..: I

1

I.

I

I

I

I

I I

I

I

0·2

•

.

I

! I i

I I I I

I!

I

I! I! I! I

os

1·5 2·0 d/nm

1-0

2·5

3·0

FIGURE 8 Effective pore size distribution for JF144. Solid lines - primary stage. Broken lines - secondary stage.

1-0 ~

6d

I

I I

I I

I I

0·6.

--.

I

I

I I I

I

i .. ....... ....., ~

I I I 1I

I

I

I I 1i

J I I I I I I I

i I

os

I

I

I

I

I

I

I

i

I

0·2

._J_._.

~._.

· · ·

··

I

I

I

I

I. - i.."'-",":'-. I 1 I

, I

I

I I

I

=."".-_'r_"I .-........ : ......, - '-'. I

....

.

. -. ...... . t'

I

I

I

I I

1

1·0

I

'1"

I

1·5 d/nm

2·0

-

i .. . . . ... I

i I I

2·5

"

··· ···

3·0

FIGURE 9 Effective pore size distribution for JF518. Solid lines - primary stage. Broken lines - secondary stage.

98

1-8

r--

-I I

I

I I I

-,

I -4" I

" .. .I

I

i .... I

I

1·0

· · ··

0'6

I I I

I I

I I

I

I

· ·

I

I

I

1-'-'-

I

I

I

0·5

I

I

I

I

I I I I

..... ....

.

.•.

- - -- . _., ._._

._._.-

L -"---';"----

I I

I

I

· I

0,2

I

··

I I

... ... ,I.,. I

I I

e .•

0

. .... \

·· 0

·· 0

3·0

2·5

2·0

1-5

'I"'

I I I

I

I

; 1-0

I 0

d/nm FIGURE 10 Effective pore size distribution for AX21. Solid lines - primary stage. Broken lines - secondary stage.

There Figures 6-10.

are

marked

differences

in

the

shapes

of

the

histograms

in

The difference between the molecular sieve carbons and those

containing a broader pore size distribution is particularly noteworthy. It is also

interesting to compare the histograms for JF518 and AX21.

Despite the

similari ty in isotherm shapes for these two adsorbents the histograms show that the pore size distributions are significantly different.

99 At present it is not possible to locate precisely the upper and lower limits of the distribution for each adsorbent. To obtain this information it would be necessary to extend the range of adsorptives to include larger and smaller molecules.

In

this manner it should be possible

to

establish

the

extreme limits for molecular sieve exclusion and secondary micropore filling, respectively. In spite of these limitations, it does seem reasonable to extend our present analysis

somewhat further by taking into account the molecular

sieving and low pressure hysteresis exhibited by Carbosieve and JF005. On this basis we have arrived at the tentative estimates of the range of effective pore size given in Table 3.

TABLE 3 Nitrogen BET apparent surface areas, total pore volumes from nitrogen as plots and distribution of effective micropore size. Carbon

Apparent Surface Area m2 g -1

Effective Micropore Volume cm 3g- 1

Effective Micropore Range nm

Distribution

Carbosieve

1179

0.43

0.3-0.7

JF005

882

0.33

0.3-0.7

JF144 JF518 Ax21

1236 1793 3393

0.55 0.98 1. 52

0.3-2 0.3-3 0.3-2

Narrow. ca. 45% < 0.6nm. Narrow. ca. 30% < O.6nm. Fairly narrow. Very broad. Broad.

CONCLUSIONS The new method of micropore size analysis described here is still in an early stage of development. In spite of its present limitations it should be capable of considerable refinement and it would appear to offer a number of advantages over other current procedures data.

Thus,

it

sets

micropore filling

out

to

take

albeit at a

for

account of

the analysis of physisorption the

different

semi-quantitative level -

mechanisms

of

and also makes

allowance for any monolayer-multilayer adsorption which may be taking place

100 outside the micropores. application

of

other

Further development of the method will

probe

molecules

and

the

study

of

involve

adsorbents

having

uniform pore structures.

REFERENCES

2 3 4 5 6 7 8 9 10 11

the

S.J. Gregg & K.S.W. Sing, Adsorption, Surface Area & Porosity, Academic, London, 2nd edition, (1982). K.S.W. Sing in D.H. Everett & R.H. Ottewill (Editors), Surface Area Determination, Butterworths, London, (1970), pp. 25-42. K.S.W. Sing in J.M. Haynes & P. Rossi-Doria (Editors), principles and Applications of Pore Structural Characterisation, J. W. Arrowsmith, Bristol, (1985), pp. 1-11. D. Atkinson, A.I. McLeod & K.S.W. Sing, J.Chim.Phys., 81 (1984) 791. P.J.M. Carrott, R.A. Roberts & K.S.W. Sing, Carbon, 25 (1987) 59. P.J.M. Carrott, R.A. Roberts & K.S.W. Sing, unpublished results. A. Capon, J.J. Freeman, A.I. Mcleod & K.S.W. Sing, Extended Abstracts, 6th London International Carbon & Graphite Conference, Society of Chemical Industry, London, (1982), pp. 154-156. P.J.M. Carrott & K.S.W. Sing - accompanying paper in this volume. P.J.M. Carrott, R.A. Roberts & K.S.W. Sing, Carbon, in press. D.H. Everett & J.C. Powl, J.Chem.Soc., Farad~ns.I., 72 (1976) 619. D. Atkinson, P.J.M. Carrott, Y. Grillet, J. Rouquerol & K.S.W. Sing in A.I. Liapis (Editor), Fundamentals of Adsorption, Engineering Foundation, New York, (1987), pp. 89-98.