A STUDY OF THE FINE STRUCTURE
OF CARBONACEOUS SOLIDS BY


       AND APPARENT
          1)ENSI'I'IES

MEASUREMENTS or( TRUE

PART 11. -CARDOSIZED COALS

13~ ROSALIND E. FRANKLIN

Received 15th Februnty, 1949

hl~~cisurenieriti of tlie true arid apparent dcnsi ties and adsorptive propertic
of coals carbonized :it 6o0-1,6jo0 C 1i;ive been uscd to study the variation 0
tlic colloidal structure with the teinper:tt lire of carbonization.  It is shown tha
the true density increases with increasing carbonization temperature, reaching

lhe porosity of particles small enough to pass a 240 n.s. sieve is large (from

ubont 2.1 g./c111.~ at I,IOO-I,~OO~ C.

I.

7 yo to z j yo) and increases with incrcasing carbonization
600" to r,ooo0 C.
tc.mperatiire between 600" and 800" C.  The acccssibili
surfaces to liquids and gases decreases with iiicreuing
tiire in the rnnge 6m9 to r,hoo" C ant1 1s governed by the
in the pore systein rather than by tlic meail diameter of pores which are uniforLn
alorlg their Ir.ngtJi.
molecular sieves, the width of the coilstrictions in the pore system being of the
same order as the diameters of tlie simple Inolecules ustd for the density

The specific surface increase with

As a result of this pore structure, the solids function

Rob YLIND E FRANKLIN 669

ementy (I e z to 5 A)  `1111s niolecular sieve stiucturr must be of im-
ce In determining both thL chiinical behaviour of the solid and the
sltlon of the gas evolved during carhomLation

aviour of colcc prepared froin any given coal 13
the niaxiiiiuni carbonization temperature In
s of rencti-i-ity nheii the temperature of carbon-
o3 C (the exact terriperxture `Lnrics fioin coal to
is reflected in inmv of the physical and chemical
tcrial To1 excirnplr, thc adsorption of carbon
to sulphriric acid, the dispersibilitv in sulphuric
erature and thc `` wrnbustibility " of a number

decrease rdpidly with iiicreailiig carbonization
C.l  In an attempt to inveitigate the undeilyiiig
nnon, Griffith and Hirst rnadc an extensive

alcohol vapours md showed that the apparent

OSIII~: to the aiinosphr~ie for 48 hr. was greatcst
about 700° C and that tlir rate of adsorption
ii e drcrcnied i,~pdly nith iciitlier iiiciea~e in

a large voliiin~ of L try fine pores and ;L large

coals whic 11 giire iio bubble-strui tuie hloreovei, the
f tlic poiosity 15 <lei clopcd at tempcratures above the
when the carhoiiizd coal hCi\ alreadv quiied a rigid
o fur thcr hiib1,le-formatioii ccm occur
ble number of meaiurernents of thc apparciit densities of

consistent with those of the present work.  This suggests tlirtt struct,lrai
changes similar to those descrit)ccl below occur during the carbonizatioli
of all coals, and are not peculiar to the specinlens studied here.

Experimental

iCTaterials .--CARBONIZED Coxrs. Samples ~vere Iprcpared from thc follo\~ing
~NTHKACITE C : S. Wales anthracite, Group AI, 94'2 yo carbon (dry,

four coals.

niiiicr;tl-free\

~J

COAL F: S. \Vales coking coal, ~lcta-bituniinous, 89.7 yo

COAL 1% : Yorkshii-e c;tlting coal, bteta-lignitoiis, 83.5 Yo cartoll
COAT. I< : Kor.tliiimberlarid weakly caking coal, hIeta-lignitous,

For carbonization at temperatures up to 1020' C. samples ground to pass a
72 B.S. sieve were 1ic;ited in nitrogcn in Kichrome-wound fiirnaces at j" C/r1lln;
and held for 2 hr. at the maximum temperature.

For carbonization at temperatures higher than 1020' C a inolvbd~:~lurn-,
wound tube furnace was used Tlie sample was fiist carbonizril at ICJOO" c.
as abovc, and, if coherent, rc-ground to 13 a 72 B.S. sieve. It was tilcIl
packed into a 4 iii.-bore carbon tuiw through which nitrogen was p:isscd.  Thc5
inolyhdenuin furnace was brought to tire desirld rliasiinuni teiiipf:raturr, and
the tube containing the sample was thcn ac!variccd through the furnace in steps
so that each px-t of the sample remained for 2 hr. in the zone of iiiaxiniu~~~
temperature.  Ihirinq the initial stages, and again at the completion of thd
proces, the closed end of a carbon sighting-tuhc  vas situated in the zon~ of
maximum tcmpeiature, and the teiiiperatui-e was measured with an oiltical
pyrome tcr.

Analyses.-- -F'roximate and ultiiii;Lte analyses of the carboniztd sainplcs
arc giv,:n in Tal,lt,s I to IV.

(di-y, Ininei-al-free).

(dry, niineral-free),

82.4 9, (c!ry, mineral-free).

'i'AB1,E I.-AKAI.YSI~ OF CARBONIZED CoAI. c

__ ~

Carboniza-
tion lrmp
  OC

l'roximatc Analysis

Adl

1 Ultimate Analysis (Parr's baaiil

itrogt'r

Liquids.-~I~'.TiihNoL, UESZENE, ACETONE, Enmi<, C~ILOROFOHRI, CAKBON

n-f-l~x~~t(: l~oiling at 67-09" was supplied by B.Il.11.

C.~RHON ~~1SULPHIl)E was purified by treatment with potassium pcrnianganate!
mercury arid mercuric sulphatc according to the inetliocl of I+aniniick
Ilowartl 6 anti was subsequently dried and distilled.

         was prcpawd by electrolysis of IO yo caustlc potasll
solution saturated ~vith baryta, and w:is purified by dilfusion through a pallzitliuni
tube.

TrrrK.&cHI.ciRm6 : Aiialar reagents v rre used.

OXYCEX, ~Ir.-rH~wje.-Cyiinder gases \YCIC used without purification.

Haminiclr and Howard, J. Chew. SOC., 1932. 2915.

Proximate Analvais

~-

Fixed
Carbon

71'1
75'3
81.7
85.2
87.1
91.1

-

Ultimate Analpis (Pan's basis)

~-

Proximatt. Analysis

I I I

RIoisture

i'olnt1:e
less
hioisture

Fixed
Carbon 1

I ~

I-- '

5'0
1.6

34")
34'0
21'9
I L'3

4.2
2.8

_.

1

Proxini,ite Analysis

I

-I-
I

672

mc'isurements of densities in Iieliuni and in liquitls have been described iil
preceding paper.'                                           Y
with time, was investigated by making ineasuremt:iits 2 hr. and 24 hr. after
                                                      Y
immersing tlie solitl in the liquic!.                              Y
continued until no further drift occurrcti in 24 hr. ;

drift in helium continued for ~iiiiiiy days am1 \vas not follo\\reil to conil>letioll,
Correction for Mineral Matter.--The density valnes recorded have in
all cases been corrected for the iniiiercil contrnt uf ihc s'inlples by uz'o ot tl,k
correction formula of Waridlcss and Macrae.8  rUtlioufli the mineral corlterlt
of carlxmized coals is soine\~hat greater than that of raw coals, the uncertainty
in the density correction is 1t:ss since the co~iipsition of the, mineral ni;l{tef
differs less from that of the ash whose density xvxs tlctcrmincd.  The corrcctiod

did not exceed 3 yo lor auy of the carbonized coals used.

FINE S'TRUC'l'URE OF CARBONACEOUS SOLIDS

Measurement of Density.-The apparatus and methods used for the'

Density " Drift ".-Tlic '' density drift," or iticrcnsc of app'ircnt densit

                In helium, measurements were nonnall
                        iii a few cases the densit

Results

APPARENT 1)ENSITIES OF C0hr.S C,\Rl3oKIZED AT 'rE>lPERATURES Up TO

rGjoo C.-Measurements of the densities of coals c, 17, 11 and K carborrizcd at
temperatures from 300-1650" C wcre inado \vi tli hclinm, methanol, water ailti
n-hexane.  With cod 11, carbon disulphide was also used and witli con1 K,

bznzciie.  Jn Fig. I to 4, densities :LTC plotted against carbonizatiorl telllPer-
aturc for tcmpcratures up to IO~O" C.  The full-l~ne curves in the FWrcs
represent clensity values after 24 hr. iinmersion in the licluic1s anc~ t~ic iid \,al~eS
obtained with helium, and the bi-olccn line curves represent apparent densities'

' Franklin, Tram Flcraday Soc., 1949, 45, 274.
8 \.\'andless and Macrae, Fuel, 1'434, 13, 4.

I

- 2.0

- If-8

-1.4

ROSA I, I K D  E.  I; R AK KI, I N

1

I 1 1-- 1

673

674 FINE STRUCTUR I;: OF CARBONACEOUS S0LIL)S

after 2-hr. immersion in liquids and the first 1-ecoi-ded values in heliuni.  The
height of the shaded areas is a measme of the density ~lrifts.  (The initial values
recorded here and elsewhere for sample-: which slioiv a large drift in hellurn
refer to measurements made as quickly a possible after admitting the gas to the
sample ; they have only qualitative sign iicance.*)

X Fleliuni. 0 Methanol. 0 \I:ater fl rz-Hexane. A Benzene.

TEXPEItATLIREs BELOW jOOO C.-fielour 500" c: the val 1d.t ion of apila
dciiritv with carbonization tenipcrat~re is differmt for e,ic!i of the four c
investig<atrd.                                         Tlie
parent dcnsities of coal F in w,Ltc>x-, of coal II in water, n-hexane arid lie
:id of coal K in water and in n-liexane are slightly decveased I)p heating to 30
arid apprtrent densities in mi~t11:tilol in all exes increase cclntinuously wit
creasing carboiii-ation t ernperarure.

For the anthracite c there is no apprcciahlr change.

trvecn 500" C and 600" C
t. c sets in allrl the densi

ly that large and signifi

600°C arid rooo"C. The

sults for coals c, H and K may be surrimarlzet~ as follows.

n-15exsrie' gives a low de~isity value which jncre;~ses continuously wi
creasing carbonization temperature and shows no appreciable drift.  The
obtained xvitli benzene fur coal K are similar.  Helinni gives a density N'
from 7 yo to 23 yo in excess of the n-hexane valtie ant1 which increases
rapidly with increasing carbonization tvmperature.  \Vheii the temp"
of treatment rises above about goo" C the density in hcliurn bcgins to stlo
drift, and at 1000~ C the drift IS large and the initial density is only sllL
excess of the density in n-hcxane.

between the sample and the thermostat was estah1ishi:d before the
was admitted.

* In all cases sullicient time elapsed to ensure that thermnl cquili

IIOSALIXD E. FRANKT,IN 67 5

e density in methanill exceeds that in helium for all carbonization tcn-
Tc.5 up to 700' C.  Between Goo" C and 750" C a density drift in methanol
p to appear and increascs with increasing carbonization teniperat~irc.
out 800" C the density in methanol (nieasiired after 24-hr. immersion)
through a maximum, and between Soo3 and gooo C it tlccrcases sharply to
e approxiinatcly equal to thc n-hcxane value.  At 1,000~ C the tlensities
ll:~nol and in whexanc are equal and show no drift.

variation of the apparent density in lvater is simi1;tr to that observed
ncthanol, except that the cori-esImntliilg changes occur about IOO~ C
Up to 800" C the density in wattr ~lio~vvs no drift sn(1 is ody slightly
han that in helium.  Fktwcen 800" and ($00' C: a density drift develops,
                                                *After
nmtion at 1,000~ C the densities of the coals c and K in water are cqual
le low, n-hexane valnes and show 110 drift, while with coal II th, density

. slightly higher than the n-hexane value and the drift is small.
rijoi~ disulphidc gives apparent densities for the c:titionizcd coal H which
-1ltical with the methanol values for carbonization temperaturcs up to
.  Between 600' and goo" C the apparent densltics in carbon disulphide
tiler higher than in methanol, although a density drift is obsei-ved in
xiinztely the same temperature range for the two liquids.  After carbon-

at 1,000~ C the apparent density in carbon disulpliidc is equal to the

11 Cod F.-The density changes which occur in the coal F in the teniper-
re range 6oo-1,ooo" C are fundamenta.lly siniLir to those descirbed above

oak c, H, and K, hut differ significantly in two respects.
-tly, the density of this coal in n-111 xane increases niore rapidly with
mation temperature, and the differ-enc(. between the densit ir:s irl hellam
hexane for any one carbonization tempcratlire is lcss ;  in methanol and
the maxima which appear in the curvcs for thi. other tlrree coals are
ed into strong inflexions.  Sccondly, all the cha~    occur about 100' C
for the coal F than for the coals c, H and K.  The clri    densitv in methanol

below GOO" C and at 800" C the density- in inc   ol has -fallen to the
ne value.  At gooo C thc density in wrater is equal to tiiat in n-hexanc.
it? drift in helium begins below 800" C and at ~,ooo' C the density in
too has fallen to the n-hexnne level a~id shows 110 (Jrift.
' results described above show clcarly that after lrt\.iting to fho0 there is,
ur coals, a large pore volunie which is accessible io hcliiim but inaccessihle

iiensity after 24-hr. immersion is at a maximuni at about 900" C;.

2-hexane value and shows no drift.

-7

-I

I

.. J

CARBOVIZATION TEWPFHAIUPE. 'C

FIG. 5.--Coal F-.

x Helium A n-Erexallc

ane.  As the carbonization temperature is irrcreasett aho~e Goo' this
umc is increased ; at the same time, liow~ ver, it becomes Icss accrssible
.Methanol, carbon clisulphide and wa ~r are successip-cly cxLlucled,

r carbonization at ~,oooO, helium too i
fid penetrates on1y slowly in the other co

676 FINE S'I'RLCTUIIF; OF CAICEIOX.ZCEOUS SOL,1 IX

The loiv values obiained with 11-hexane for the coals carbonized above boon,

are sharply defined and the apparent d
a real characteristic of the solid mate

                   ty in this liquid appears to rcpres~n
                    There is no appreciable drift, an
                  01 and carbon clisulphide fall to th
                  On the othcr hand, apparent densities
               helium values are associated with large
drifts, showing that penetration of the pores bq- the liquids is slow and inconlpl,te

  RANGE 1,000~ C; to 1,650' C.-The general pattern of the results outli
above is further clarified by measurenients made on sariiples cuboiiized
teinpcratnres above 1,000'' C.  Results for coals I' and H are shomm in l?i,
and 0.

                                       hj
For each of thc three coals c, F and H carb,,nized at l,OooO, the denjlties

I 1.

500 7w I000 1250 lsoo

CARBONIZATION TEMPERATURE ,'C

IZIC G.-Cod li.

2: Heliuni. k> n-I-lexane.

given by nietlinnol, Tvater and ?z-hcxari, :ire FcIiia1 and show no drift, and this
is true alio of higher carboniLatlon terllp:1-atures.
in helium, too, coiitinucs eqi~il to tlie \ralue (,htained with the liquids.

For tlic coal P, the tlci
               T

i teni1~eratur-c from I,OOO~ C, to r,6.50°
id coal I' in hcliiitri ani1 in liquids incr

The density of the carbonized co;~l 11, ineasurcd in mcthaiiol, \vater or 12-
liexane, incrcaaes only slowly above I,OOO', reaching ~'75 g./c111.~ at 1,600' C
The density in hcliuin at first increases, anti tlien decreases to thc Ion value
given by the liquids.*  The highcst densities in lielium are associ,itcd with large
arid prolonged drifts.  'i'he nlaximurii YaIue recorded was 2'03 g./~rn.~ io
carbonization teinperatiirc of 1,100- I, I 30' 1iut siiicc the drift was Ilot foil<)\
to conlpktion, this figure is lower than the true density of the snnlpl? ; a 1116.
value was olitaiiietl when the sample W;LS nore finely ground (sce below).

The Influericc of Particle Size on Apparent L)ensity.-All the res1il.k
described above were obraiiied \\.it11 s;iniplej grountl to pass a 72 1I.S. sieve.
Mensuremetits ivc1-e also 111;ide 011 smiples of co;~   carbonized at terI1perature~
above 600' C mid ground to pass a 210 B.S. SIC   .  The results are .iho\m in
Tabie V.  Tlie dcnsity 111 ~iictl~a~iol of the sa111pl  arboriized at Soj" C and the
densities in ~ieliu~n of tilose carbonize(! at I, loo0 and I,2jcJ" c were consider:.l+

* The density in helium of the anthracite c carbonized at 1,4jo@ C ~holvs no
drift and 1s equal to thc n-hexane value (oht;tiried by intcrpolatiot~) for cod Ii
carbonized at the iame temperature.  SlJice t1ie apparent densities of t
anthracite c and of tht. coal H carbonized hetv~een 700~ and 1,ooo" C and measllr
in helium, watt-r and whexane are almost identical (see Pig. I and 3), it is Pro''-
able that tlie density of coal II in helium after carbonization at 1,450' c \\
be equal to the low tz-hexane value ani1 would have no drift.  This redt
takcn into account 1.1 drawing the broken part of the helium curvc in F@ ''

lIOS-41,IND E. FRA?U'KLIN 677

increased by reducing the particle size.  On the other hand, the resulrs obtained
%,iell n.hexane for sanlpies carbonized above 800' C and with helium for those
                      Thus, where no density drift was
rved the density was not increased by grinding to the lower size limit.
density drift in helium of the samplc carbonized at 1,000' C was completed
  nsiderably less than 24 hr., and in this case, too, the density measured after
                       l'or samples which showed large
      density drifts, the apparciit density after a given time was highly

below 1,ooo' C were not altered.

hr, \vas independent of the particle size.
3 pendent on the particle size of the material.

TABLE

V.-~FLUBNCE OF k'AKTICL*> SlZZ ON r\PPAlIENT IhXSITY OF

(;ARIIONIZED <:OAT2 €1

Size

72 H.S.

72 i3.S.
240 1'1,s.
72 E.S.

72 U.S.

240 r3.s.

240 1i.s.

240 13,s.
72 B.S.

240 13,s.

PI .:lium

Drift,

Yo

The maxiniiini density valuc recorded \vas ~.(ii) g./cni.3 for coal H c;trbonized
    The true density of the material cxcr;~xls this value, since ii slight

iooO C.
was still observable aftcr 'j clay~s.

Influence of Riolecular Size.--The order in w-hicli n-hexane, nit.tliano1,
   licliunii are eucluilcd froin the pims ;is I lie (.,irhonization ternper;tture
   - the s;init: for each of tile four wiclc~l)~ different coals c, F, ii ~i~itl K.

tl I e ord er of decrc as i ng molccul r~~ si zi: .  Furt1it.r nieasurenir:n t s
tlic carbonized coal I{ with accioriz, rther, c,hlorr-iform anti carbon
tliesc, together with thr resiiits olitaiiicd with cai-hu~i disulpliidc

   I :mtd with benzcne for cor11 K ~CJIifirlii tlia t rriolccular size is the principal
icictor determining the po.,ver of organic liq111~1.; to pciietratc into the pores of
        The results obtxinctl with 8 liquids :ind with hchum for cod
  onized at tenilpratiircs betweeri 7m0 ant1 I ,000' C are givcri iri Table VI,
  ier with tlic molecular volumcs of the liquids at 2.j' C.  The 4 liqiiids
  .have the largest niolecular volunici give appareii t densities approxinintcly
   to the low, .n-hexane values.  Iio apparent dciisities appreciably lover

'chn tile n-hexxiie va~ues have been observed.

Adsorption of Gases on Carbonized Coal .--The density of he-structured
c*ybonaceous solids cannot be measured in gases other than helium si~~ce all
ohr gases are to sonie extent atisorbed ature. Howevcr, since

ts of the apparent rate
of penetration of gases
, the roon-temperature

of hydroKen, methane and oxygen on coal H carbonized at Goo",
,ooo' C briefly investigated. 'lhc apparatus and method M ere
as those used for the nieasureinents of densities with Iicliuri~. The
e of the sample was first determined with helium, which was then
, and the measurements repeated with one of the other gases.  The
etwveeii the apparent v,ilumes of the sample in this gas and in hclliuin

als.

adsorption is a very rapid proc
ption may serve to study qualit
fine poi-cs in carbonized coals.

gave

vo~ume of the gas adsorbed.

'4 *

€Ieliuni . .

Water .

hIethaiio1

Carbon di-

sulphide .

Acetone

Chloro foiin

Carbon Letr,i-

chloride .

Ether ,

rkIisity, g./c111.~

Lknsity aftcr 2'1
hr,, g/cin."
Drift, "/; (2 hr.
to z,+ lrr.)
Density aftrr 24
hr., g./~rn.~
Drift, yo (2 lir.
to 24 hr.)
Density after 24
hr., g./cin.3
Drift, (2 hr.
to 24 hr.)
Density after 24
hr., g./cn~.~
ljrift, yo (L hr.
to '4 hr.)
Dcnqity after 24
1 ,I-. , g . /.In, 3
Drift, (2 lir.
to 24 lir )
Ilcnsity after 2'1
hr. g./c111.~
Drift, yo (2 hr.
to 24 hr.)
Iknsity aftei- 24

]'rift, 9.; (2 111..

Jlcrisity after 2.1
lir., g,,'c~ri.:~
Drift, O/(j (2 lit
t(J 24 1ii.j

DZlft, 9;

111.. , g , /LIIl.R

t<J 24 hC.)

7joo c

100" c

* Obtaiiicxl by interipolatioii

I~YDROG~%N :

       'The qiiantities of Iiytlr-ogcn aclqorlxxl on the three samples
are slionx grapliically in Fig. 7. 'J'tie hvat of arlsor pt ion, calculated from
nie:isuremrnts msde at 1s' anid 2.j" C, \vas about 1,800 , nl./g.-mol.  Since the
isotherin.; s!.riwn in Fig, 7 are approximately lincar, and the heat of adsorptioil
is tlie same for each sample, thc SlolJt'S of the isotherms give a rcl a t Ive Irieasuie
of the specific surfaces.  Retucen Goo" and 800" C the slbecific surhce iiicreases
by about 40 yo.

Xdsorption cquiiihrium was estnhlisIiecI iniinediately (i.e. in less than 4 niin
the time reqiiired io! making the first nleaiuremcnt) on the samp1t:s carbon
i7ed at Goo' and 800" C.  On tlie 1,000" C sample, a iapid initial adjorpti
fc,!lowecl l)y a large adsorption diift, and equtlitxium was established
2.t hr.  I{etween 800" arid r,oooc there is a small dccvcase in speciiic s
arid the gmd agreemcnt between tlie resnlts obtained with the 72 m-'
thz 240 mesh sample suggests that this dccrcasc IS rcal.

           lie results obtained wit11 mettiane are givcn in 'l'n~~~e
The lieat of adsorption on the sample carboniLed ;it 6ro", calcukiteC1 frorn
nieasurcnietits macle bctwecn 12' and 30'. u-as 5,800 cal./g.-inol.

Much larger quantities of methane xvcre adsor-hed than of hydrogen
the proccss was considcrably slower.  \Vith the sanllile carbonized at 61
there was a large initial atisoi-ption, $1 yo of the total occurring in the

hrKTHaXE :

eqililihriuin was established within 5 hr.

                   011 the 810" C sample less
the total adsorption occurred in the first z min., and 5 days were

,xn;-;.o/;

[z H CARBONISCD AT l00%'$20s5
10 COAL H CARBONliED AT aOS% 02 85 5
~

COAL H CARBOHISED AT lkOOoc (72 B 5 5

A'

7T -r

I

I I I I

in 'Table 1-111

liquids, the larger the niolecule of the gas, the lower the carbonization teillper-
ature for which it is first excluded iron: the pores.  'The rrsults arc summarized

TABLE VI11

C. as

IIeliiiin .

Ilyclrogen

Oxygen .

Me thi 11 e

600" c

Xo drift, pi'ne-
ration comjiletc

h-o drift, pene-
.ration corripletc
and adsorption

rapid

Larae rapid ail-
,orption follo\vvtl
hy long slov:

r,f total atl-

crirrwi in first
   z miri.
Ail Y o rp t ion ei 1 ~i i-
liljriuni eitnb-
li4w.l in j iir.

zidiorption oc-
cuine(l in first

dl-ift. SL y;

b( I sp ti 011 oc -

h J 94, Of total

z iiiiii.

iXo drift, pciie-
rntioii complete

No drift, pile-
.ration complete
and adsorption

mpid

A? for Gooo C
b1nt rather
slower. j j 76
of totnl :Hi-
sorption oc-
curird in first

2 ll1lIl.

Ails or pi i un c qui -
liI,iium praLtic-
all> estaiilislied

ROSALIND E. FRANI<I,IN (18 I

pvre: 5yste111 : ' after carbonization, the fine-structure porosity, surface area
g6t1[.,accessibility of the pores are all less than for tlie other coals carbonized
%#&e temperature.

Flti. s

e-Structure Porosity of Coals Carbonized above 600" C.--Tt
*

om the results described above that the low apparent dc
carbonized coals with n-liexu;i.~ie or with other liquitis of
is substantially indepcrtdent of the espcrimerital condit
ned and sigriificniit property of the solid?. Tlicrc is
iat it is approxim;itcly cquJ t   le lumli density of the iiiclivi<lii:il
Not only is it constant for a u   raiigc of liq~~icls, but it \vas also

-1IP

at room teniprratui-e f(>r I i !. ;liter e  uation arid hcfor-i. acl:n1ttiIig
The prewiice of air would otiviously iniIicde the diff1is;oii of a liqiiiii
mi-t:s, and tlie s;inic treatii~r:rit did, in fa( t, reduce considerably the
cleiisity of the s:mie sarnplt,s in 1net11;inol. LIorcover, hI;iggs 11~1s
                                     Tli- amount
satnrativii pi-cssure at 2.j' C corresponds approximntcly to that

rm a monolayer 011 thi: external siirface CII the powder.

ay lie considered to 1ne:iiiire thc. true density of carlmnizeci ci~al.:
rbonization temperature is riot so high that tlie gas is excliidcd floni
r a part of tlie pore space.  This conditian IS inlfillcd for i.oal> r,

rboiiized at temperatures up to (poo C, and ff.)r roal F at teriiperaturi,s
  It folloivs thctt tIi\; fine-struc.ture pui-clsity of tlicsc inateiials is

when %rmplcs c,i ct-d R P''

e adsorption of n-hexane on coal K carhoniied at Goo".

? C.

P = (dHe - dhex)/41e,

and dhex are the dcnsities in kwliuin and n-hexane respectively.

9,' P is piottcci against carI>onizc+tion terriperature.  It is seen ttiat
ut the range of carbonization temperature for wliich thi true density
le (by helium) the fine-structure iiorosity increases urith increasing

fiile-structure porosity increases wli~le the acccssihility of tlie
.  This 1e;xds to the conclusion tht the accessibility of the pores
' the size and frequency of fine constrictions or " bottle-nccks "
tern, and not by the mean t11,tnietrr of pores which are u~iifonii
@h.  The alternative assumption of unifolm pores would require
casing temperatiire, the dianieter of thr Ipres steadily decreased

r all fonr coals.

682 FINE STRUC'IUICE OF CARBONhC'II0'IJ.S SOI,IUS

while their number increased r;tpidly, new pores formed at any one tempt.rature
being narrower than those formed at a lo~x-er temperature.  In such a system
the increase in porosity would be accompanied by a still larger increase in specific
surface.            er, that between 800" a11tl I,OOO' C the spc,cific
surface decreases while the purosity increases.

Tilc striictural changes which accompany increasing car1)onization temper-
ature between 600' and 1,000" are associated with tlie 10s.: of volatile tlecom-
position products froin a rigid solid. 'Thc existence of the clccomposition products
(mainly CO, CII, and H,) implies that there is a re-arrangement of sur-face atoms
or small group;; of atoms within the solid.  This nlay be su:ficient to  account

It has been shown, 1x1

0 Coal c. 1:: Coal 17. Q Coal 11, x Coal I<

for the creation of a niore lij@iy
constricted pore system resultlrlg
from iiicrtascd interriiicellar con-
tacts.  A similar effcLt iiiay per-
Iiay~s account for tlie slight de-
crease in surface area between
600' and 1,ooo0 by eli~niiiation of
surface roughess. The increase
in fine-structure porosity, on the
other hand, shows that the rigitl.
it>- of the structure is too great
(ur the carbonization temlxrature
too low) to perinit such large-
scale rc.-arraiigcmcrit as wo~ld 1ae
iiecesxary to compelrsnte for the
loss of volatile matter and the
increase in true density.

>lolecular Sieve Properties
of (hrbonized Coals.--Thp
\olunie of fine pores in coals
carbonized at I,OOO" C or at
Liglicr tcniperzitures niay arnount
to more than 20 y:, of the volume
of the solid. 1;or carbonization
t f: nip era tu 1-r s bcs twccn 6ooo and
I ,o0oo (; ti.e solitls i~i,liavc as
niolccnlar :IICV~S ; in any given
s,i:iiple dl iiiolecules lxi-g~,r than
;L ct,rtain s,ixe .LTC c,oiriplctcly
psclii lctl froni tt,c pores. Car-
I,01117d coals ni;~y hi. cijinpnrcti

pcc,t with thc zeoiitcs
lktrrvr,M altiiougli tlieu
cai);xity is 111111~11 less