HEAT OF REACTION OF
DOUGLAS BRUCE SMITH
A DISSEP.TATION PRESENTED TO THE GILDULATE COUNOL OF
THE I'NIVEPSiln OF FLORIDA
IN PARTIAL FULFILLMENT OF THI REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
AC Y-R:.LE DGMHET S
The author wishes to express his sinLc-r appreciation to
Dr. Herbert E. Schwevcr, whosc interest, advice, and criticism
Stimulated and guided this research program. H. also wishes to
thank Mr. John k'. Kalway for his assistance in the conLsructLon
and maintenance of the equipment, Mr. D. S. Smith ano Mr. D. A.
Falgout for their assistance in the laborarory, his Ilfe for h-r
encouragement and patience, and the members of his Suprr', iory
Connmittee, Dr. Mack Tyner, Dr. F. P. May, Dr. W. S. Brey, and
Dr. J. W. Gaddum. He is also indebted to Dr. H. A. Hoi,'ur, thu.
until his death, proviaod guidance of the author's minor program.
The assistance of the Carpco Enginc-ring Company in its
la IIIIII f dthe high voltage rectifier, the Te.xas Company and the
iiiiii Shiii ll Oi i CopIIIny in supplying the asphalt, and the National
IIIIIIIScifI |||ii Foundation in the financial support of the research
TAELF OF CONTEilT3
ACr .'LLF L J LLUilTI . . .. . . 1
LIST uF TLELES
LIST OF FITlG ES . . . . . . "
ABTRAC. . . . . .111
I. I .:JilkOjDUC[TI[OL ....... . ...... I
II. THE'i ......... ..... . .
A. A.phjl Co ,p: i o . .... . .
E. FraiCLtinaLion o[ A-ph.iLL ... . . ... .
C. Ir.,l.lionalilon b, Chriuma3LO raph . . .
D. Fr.ctionirion b Sol,.',nc LExtrac ion . 11
E. Asplihal R.Jc ions . . . . . .. 12
F TheL Alt Llo.ning K-iaction . . . . . 13
G. Cacalytic Air Blau'in; . .. . 16
H. The Sulphur KR ti, cion . . . . . 17
I. lih H ac oi t.i Lion . . . . 19
II F..PEPRIJ[IJTAL . . . . . . 22
A. Drsign to the Pruc s. . . .... 22
B. D.sign of EquLpmtnnc . .... . . . 23
C. Exptrir'encal Disigin for Air Bli.ing . 29
D. Operating ProcL-dur, fur Air Lloing . .. 32
F. E.xprimmental Design ior Catal/cti Air
Blowing ... . . . . . 36
F. E:.-plrirn L.tal Design [or Sulphurization . 37
IV. CALCULATION PROCEDURE . . . . . ... 39
A. The Heat Balance . . . . . . . 39
B. E'3luation of InpuLi In1d OLtpuC TImcr . .. 40
C. Evaluation of Accumulation Termas .. .. 41
D. Solution of UnstL-ady State Huat
Transfer Prublci . . . . . . . 43
E. Evaluation of Hcac Lossc ....... 46
F. CalculaLion of Oxygen Consumed ....... 47
TALLER OF COHTIEHIS (Conriraud)
*tJ S LTS
A. Air El'.n. in;
B. iaralyric Air Llu'. ing
C. Sulph ri zati n . . .
VI. ANALYSIS OF P IJULTS .
A. Air ic..irn .
P. l-rpoiar ur, Errfc r .
C. Caralycic Air Blu,.in .
D. Sulphurizatoin of Akphalc
VII. CO CLUSIOIS . . . . . . . .
APPENDICES....... . . ..
. . .
A. Dri ingE of Etlulpi.nr . .
B. C.lciul Ialan :' Boundar,, Conui iri n .
C. Stability of rhe Finite e ift r o'nL Equia3 un
D. The FORTPAN Pro ra .. ... .
Calibra onS . .
CalLul 3ion ou H3ets
StatlEiLaL DtC .
Tabuliatcd D.at .
r F'-: acrir.r. . .
LIST OF REFERJ'CES ..... .....
BIOGRAPHICAL STCH . . . .
LISi OF TABLES
TAPLF. P :-
I. ASPHALT PROPERTIES . 30
2. HILAT OF REACr[ION OF AIR BLOWING r.SPLHLT,
TEClMP-K'\-ljR -t.0 .75 'F. . . . 9
3. HAI A OF RLACTIO; OFIR BLOuNl.G AS?IlALT,
TFiPERATURE 535 55 550 ':F . ... . 50
-. AiALYSIb OF VARIANCE OF ASPHALT AIR
PLOWI, RESULTS . . . . 51
5. HEAT OF PJ.ACTION; OF C.'TALYTIC AIR BLU0lI1IG
ASPALT . .. .... . . . 53
u. GREAT OF REACTION OF SULPHlURiZINl ASPI-,,L .. 55
7. ANALYSIS OF VARIANl.CE OF ASPHALT
SULPIIUPJRZAtIOl, RESULT . ........ 5.
SUMMARY OF AIR BLOWIIG RESULTS . . . .. 5. .
9. CALCULATED tLAT OF PREACTIOIC VAJ.UES FOIR PROPOSED
AIR BE iUlJIG F.ACTIO;;S. . . ... .. 59
10. SUMAR Y UF CATALYTIC AIR bLOWII:G RESULTS .. 0
11 SUMMATlRY OF SULPHURIZATION RESLILTS . .. 67
12. CALCULATED HEAT OF REACTION VALUES FOR
PROPOSED SULPHUIPZANON REACTIONS . . t
13. HEATr RANSMISSIOIi LDAA OF THLERMOEESTOS
I ISUIL.kIO[ I . . . . .. ... . .. S.
14. CRECK OF CALCULATIONS FROM VALUES OBTAINED
WITH HEATER . . . . . . . ... 95
15. REACTOR HEAT LOSS VALUES . . .. . . 95
LIST OF TABLES (Contrinu d)
lb. IOI:YCEN ANALYZER FJ.ADINGS L/ITH PURE, DR'i AIR 90
17 EXCERPT FROtl E.-cERIIMmTAL DATA, RUI. R-Dj-, . LO[
Ib. EVALUATION OF INPUT A!'D OUTPUT TEMiIS .. . 107
1'. TElIPEP\ATL'U AS A FU.iTION O.F RADIUS . . LOo
20. UlRIEXJCAL IlliEGRATION .. ......... 09
21. E.-CERPT FROM E.lPERIlfElJTAL DATA, RIUN R-t-3-11. 112
22. E'.VALL.ATiOII OF IlIPUT AID 'L'TPUT TERlMS . .113
23. TEMPEPRAURF As A lNl'CTIOTJ I. F P.F-IUS . . 1-
24. NUMERICAL IN;TECGRTIO .; .. .. . . .1
25. SULPIURIZATIOII STATISTICAL DATA .. 117
2b. CATALYTIC AIR ELBOWING STAiISTICAL DATA . .. 118
27. CATALYTIC AIR BLOI:ING Ai;ALYSIS OF VARIANCE .. 116
2?. AIR BLOl.ING FACTORIAL DESIGN; DATA . 11
29. AIR BLOWIGC DATA . . ... . . 1
30. CATALYTIC AIR BLOWING DATA . . .. . . I
31. SULPHURIZATION DATA . . . . . .I
32. OXYGEN ANALYSES OF SAMPLES .. .... .. 1"
LIST LF FIGUI.RE
1. E:prim,.n nt-l Proc .aiing EquLpmnt . . 2.
2. SchmTiai ic Diagram of Arpp irat, s . . 2.
3. Title-T.ip. .ra-ure Pro'f ll i = fiL), Run R-u3-o 35.
E. xp. ri ritual R. acror . . . . 78
5. Reactor Cov'.'.r . .. . . . . .
c0. EL~.c rr'. at ic Pr c ipL tat r . . . 80
7. Double PLp.- !I' L ExcharL,. r . . . . 8il
8. Picked Col.u in . . . . .2
9. Dring C i umn . . . . . 83
10. G Pr h. r . . . . . . 84
11. P r lihea t Bi l . . . . . 85
12. Pru:he3Lor CO.'r . . . . . .
13. E:.ic FIl..T [Er C lr Cah braur'. CU L . 97
Abstract of Dissertation Presented to the Graduate Council in
Partial Fulfillment of the Requirements for thL
Degree of Doctor of Philosophy
HEAT OF REACTION OF PROCESSING ASPHALT
Douglas Bruce Smith
Chairman: Dr. H. E. Schweyer
Major Department: Chemical Engineering
Experimental procedures were developed and experimental
equipment uas constructed for the purpose of measuring the heat of
reaction of asphalt air blowing, catalytic air blowing, and
,,.. the air blowing process, three asphalt of varied
conipoittLon were studied at two reaction temperatures and three
sofitenin;g point ranges. It was found that the reaction was
IIanothrmic I I ith I||a beat of reaction varying between -61 and
It|-72 ,li calories per gram mole of oxygen reacting depending upon
IthIIe IIIIII p tiollin eQf the asphaLtiiiiand the reaction temperature,
vkiteh were ishowa tolbe statistically significant factors. The
Iextlent E c Wllllllll eton of the reaction had no effect on the heat
of reaction EkydroEenation of naphthenic structures was pro-
pold to bethel pfpncpAl1 reaction occurring, especially in
AS((lg ha hihi nahe he) n.ic or naraffinic content.
mi EE im m:Miiniiiinuii: iinuiiinuiiiii nui lliiinlimin
Polymerization and the formation of oxygen containing groups also
contribute to the reaction. The effect of temperature on the hest
of reaction was shown to bc more pronounced uith asphalts having a
high aromatic content.
Addition of phosphorus pentoxide to the asphalt proved to
ha.'e no effect on the heat of reaction, but addition of aluminum
chloride decreased the magnitude of the heat of reaction by 9 kilo-
calories per gram mole of oxygen reacted, a difference that was
statistically highly significant. Aluminum chloride was postulated
to retard preferentially dehydrogenation reactions in favor of
reactions forming oxygen containing groups.
The sulphurization of asphalt was found to be an endothermic
reaction with the heat of reaction varying between 8 and 12 kilo-
calories per gram mole of sulphur reacted depending upon the type
of asphalt, which was shown to have a highly significant effect.
It was concluded that the naphchenic rings in the asphalt are
completely dehydrogenated to aromatic. Polymerization by both
sulphur and carbon linkage was indicated.
Both sulphur and oxygen were postulated to dehydrogenate an d
polymerize asphalt. The primary differences tiiiille two were that
oxygen dehydrogenated only one or possibly twuoi.s of the
naphthenic rings, while sulphur dehydrag mfled all three bus
and that polymeria ttion is ano hINiLtsti LL e by phr lIna i
the sulphurizatteno process thtBl iby Ieln lise a. r Ai
I NT kR uUCTIfON
Asphalt is one of the oldest enginevering materials known
to man; ics earliest recorded use was about 380u B.C, There arc
records of the use of asphalt by thi- HfesopLot:1n ans, Babyloni ar.s,
Assyrbans, Greeks, and Romans The comparative-ly rLcenr expansion
of highways has resulted in the us- of asphalt in cremn.ndcusly Ilrge
quantictts; in 1962 over 200 million cons ot asphalt -'er.: ustE iin the
United States alone. Since 185-, when the first asphalt roadway was
laid in P.ris, paving products have accouncdri for most of the asphilt
produced. At prt-snt 73 per cLnt of the tocal production is usnd in
roads and airports. Roofing products accourct for 17 per cent wich
the remaining 10 per ce-n finding application in such miscellaneous
products as tiles, floor coverings, paints, iiusulation, rust preventa-
tives, and bituminiztd paper products (I).*
Asphalt is frequently processed chemically to modify its
physical properties. The most widely used of these processes is chc
"air blowing process," which was first patented by C. J. D'Srmedt (2J)
in 1881 DeSmedt found that oxidition of asphalt resulted in a
*Underlined numbers in parentheses refer to the Lcst of
References at the end of this dissertation.
product hhich had a great. r tenacity and eas less brittle ad 1iss
liable to be. afiected b:: air or I:jter. In [8'.. F. X. i .erley .15)
dc.'loped a proc-es for o.idizing asphalt which consisted of blowing
air through asphaltic oils maintainLd at temperatures between .00 and
cOj OF. The present practicL as reported by Abrahai \k) is tc blo-
the asphalt at a ctemprature of -50 to 375 OF. at a rate or J0 to
50 cubic feet of air per minute per ton of asphalt for a period of
5 to 12 hours. The reaction is carried out industrially *zih' h r as a
batch or continuous process.
A modJtication of the air blowing process is the production of
"ctailytic asphalt" in Lhich asphalt is air blo.:n in the presence of
a catalyst The advantage of catalytic asphalt is a modiftcatcln of
the physical propertiLs whichh is desirable for certain applications.
The first patent for such a process was issued in 1899 to
J U. Ha'yard (.01, who recommended the addition of Itmestone dust
in the inlct air. Since then there have been many different
catalsts reconmcended and patented but the ones most widely in use at
the present are the metal chlorides, phosphorus penEtoxde, and
A third method of processing asphalt is the treatment of
asphalt -ith sulphur. Although this process antedates the two
previously mentioned, being first patented by A. C. Day (19) in
L86b, it is no longer used industrially, having been replaced by the
air blowing process, which is cheaper and yields a similar product.
In this proc-,E, knuon aJ rh'- "Dubb' Prsices" afcer .j subSIqucnI
ptc.ic t holder r (2), 20 ro 25 per ccnr sulphur as rtdctLd i'th isphllt
at 3 C liemperjrure of a bbout. oUJ 'F. uncil ch._ ~v,-.l,,r.i.n .of aJa c ai cd
Although this process has not been used in nearly: .L0 .,',ars. Int. rest
in it has bLcn re-v;rFd because of ch work of Uif'ori anj
Vlugrer ( 1 5) .
A. Asphalt Conposi t i.n
AsphalE is usually regarded as a coraplex mi:.ture of high
molecular weight hdrocarbons. A mor- specific statcm.nt concerning
chcdlcal composition is difficult becausi- thL cxtrmneL> ldrge
number of compounds occurring in asphalt make chemical analysis
e'ctLdingl:, complicacLd. The classiftLacion of asphalt as a hydro-
carbon is itself fallacious since varying quantities of o:.:gen,
sulphur, and nitrogen are kno.n to be present in asphalt molecules.
According to Labout '57) and Kalichcvsk:- and Stagncr (5_), the
oxygen is present in the foinr of carboyal. cjrboxylic. and hydroxyl
compounds; nicrogen is present as porphyrtns and porphyrin meal
complexes; 'hill the sulphur, as .-cll as some of the nitrogen and
ozx,gun. appears in hetcrocyclic rings. Hempel (41). Sergienko and
his co,.orkers (9o), and Abraham (2) have also reportedL the existence
ot nickel, vanadium, and traces of other metals in asphalr_.
The molecular weight of asphalt molecules has been reported
by Pfeiffer and Saal (80) and Marcusson (LO.) to be in tht range of
400 'o as high as 100.000. The more recent evidence of W'inniford
(lt)), Eckert and Ureetman (24), dnd Criffin, Simpson, and
Miles (35) indicates, however, that the upper limit of
molecular weight may be no higher than 4,000 or 5,000. The
high values previously reported for the molecular -eight of
asphaltenes, the highe'sr mol-cular weight fraction of
asphalt, ar- actribut-d to the e:-lscence of frret radicals in the
asphaltene molecules which form low energy bonds beti-een s..vcral
asphalten miolLcules. The ExistCnce of free radicals in asphalt
has bcen rcportid by Leybourne and Schw.'.yer (58), Y--n, Frdman, and
Saraceno (110), Brown, Cutowsky, and 'Jan Holde i12), and Pitchford
(81) using different methods.
The structure of the asphalt molecule rI rjins pr.cty much
unknown. Evidence to date indicatc-s thac an "average" asphalt
molecule consists of a condtinsed ring Structure containing either
naphchunic or aromatic rings, or both, with paraffinic side chains
attached to the ring. About half the carbon ators ar- r.:port.*d by
Murphy (71) to be in aromatic or naphthenic rings, the oth. r half
in side chains. The mass spectrometric analysis of asphalt by
Clerc and O'Neal (b1) indicates that most of the carbon is in
aromatic or heterocyclic rings. Lichmanri (112) has found that ring
clusters are connected by aliphatic side chains. According to
Alexander and Shurden (3), olefin structure is normally present
only in asphalts derived from thermal cracking operations.
Labour (57) has stated that the components of asphalt may be
considered to contain groups of four basic fomns: Saturated aliphatic
groups or paraffin, naphthenic groups, aromatic groups, and aliphatic
groups with olefinic double bonds. All four basic structures nmay
exist in one asphalt molecule.
.: . .....
B. Fractionation of Asphalt
Sinci- sph.ilts from. different sources are knoun to vary in
their composition, it is desirable to classify, specific asphalts
according to their general chemical cOmposition. The litirature
shows numerous srparacion proIcedur-s which physicall, separate
asphalt into comnpon.nts arbitrarily designacec by the individual
The oldest and one ot the most widely iccepced classifica-
tLons is the method Jdeelopcd by Marcussan ..bJ) in i91o, which
designates the asphair Licmponent. as asphaltenes, resins, and oils.
1Th1 asphdltents are Ehe high molecular weight fraction obtained by
precipLcttion methods using non-polar hydrocarbon sole.,nts. The
resins are the intermcediatc mrolecular *cight fraction thich is
adsorbed on variouss solid adsorbants The 'r.Lttion not retained on
the adsorbani is called th- oils; it is the lightest portion of the
asphalt and similar to lubricating oil.
The asphaltents are considered to be cyclic in structure,
the rings resulting from copolymerization. The rings are reported
to be naphthtnit as c.ll as aromatic uith attached side chains three
or four carbon atoms long. Barth (S) reports that the asphaltenes
contain the greatest percentage of oxygen, nitrogen, and sulphur.
The carbon-hydrogen ratio is rEported to be as lou as 0.8 and as
high as 1.0 by Bestougeft and Eargman (9), Griffin, Simpson, and
tlIleS (35), Alexanlan and Louis ( ), -iin ThurEtun an.! ni,.ul li2)
inv'stigatiiig different asphalts.
The asphaltene fraction has becn subjected to infra-red
analysis tb, Y nii and ErdJm.in (10t) and x-r diffrac c t t.,' 'i'.n,
Erdm.an. anJ Pollack (109). Infra-re-I analYsis sho-s the irc.natri
clusc..rs co be p rit-ci nd rincd thili th nraphchenict ciu cEras arc
both p, ri- 3ld ksca-condenseri Ihcat trcntatent of the ,aiph il enc
results Ln somt kata-condJn- atiun of iriomatics. Thi i:'.itcnc: ..f
opel .hain carbon/1 groups is .'etrfiei d, and p.ssibi,, aryl kc':,n s
lMany of the aromatic cluastrs are .ubscltuted to a hig.ih *aigr : b
naphch.nic and paraffinic groups Tht aliphatic portion of the
asphal ctenL structure ,as assumed rt consist of sh-,t, unbranch~_d
chains. Incertstingly, as the sire of the clusters increases, the
number of terminal sid t chains dt.cr.asi This findi'Ln is consist nt
with the belci f that within lthe asphialcic fraction of 3 siogle, cruje,
the average siL. of the arornatic clusters increase s with mol-.ular
weight and a larger number of aliplatii sitd chains are rcquLreJ to
link the clusters tugEthEr.
:-ray diftrattlon data ilndiLcatE that thi asphl-iln. mvolrculcS
consist of clusters of aromatic rings condensed in flar. shirts con-
nected by saturactd carbon chains or a loose net of naphthcnic rings.
Ergun and Tiensuu (J7) believe that it is not unlikely that hccru-
atoms replace some of the carbons in thi shlrrts and that somin of the
rings may be partially saturated. i'n, Erdman, ani Pullack (L'9)
state that if the asphaltic polymer consists of islands of con. onsed
aromatic rings cited tog.:chhtr ich chains or saturaced rings, Lih.n as
th, islands become sm.ialler, rhe proportion of saturated struc r-rs
around Lh,: aroliTLc Lslands might reasonably be expected to incres...
Emiprical formriulja for ti asphultenes of four asphalts
have bcLn reported b, Griffinr, Siimpson, and Miles (3i) and b.
Bostougeff and Barrnan kO), ihoc postulate ith- Ll:istencc ot four or
five mcmbered rings. Structures of asphaltcne nolccules have also
been dtscribed by Hillman and Barnett (42) and by Hurphy (71).
B-rch (S) reports that the resins fraction is darP-col.ared,
heivy, very adhesive, and solid or semi-solid, although LL bcc..,mes
ver., fluij or heating. The molecular weight of the resins varies
from abouL 600 Lo 2000. Resins conJain sone. sulphur, nitrogen, and
:..ygen. Knoel.s ec -a. (5t), Bescougtff and bargman (9), and Hill-
man and Barnect (J2) have found that the carbon-hydrogen ratio !varies
trom U.5 to 0.65 dpcpnding on chL source of the asphalt. Polycon-
dEnsed rings are videnc in rtsins 'ith Longer or more numerous
alkyl side chains than the ssphaltcnes since less than 35 per cent
of che carbon is in ring form. Ihe resins contain proportionately
more saturated than aromatic structure according to Barth (8).
Hypothetical structures for resin molecules have been proposed by
Cardner and his coLorkers (32) that exhibit chain branching of
alkyls accached to rings, which could form a large molecular inter-
locking network, spatially equivalent to cyclization. A proposed
chain structure for resins and asphaltenes consists of polycyclic
nuclei linked Logcther by carbon atoms or possibly by oxygen or
sulphur atoms (42).
The oils fraction is white in color and majy, in general, be
considered to consist of a c xture of saturated hydrocarbons in
the molecular weight range of 360 to 800 (8). They are composed of
single or condensed naphthenic ring (cto to five rings per molecule)
with several side chains of varying lengths, although Rossini (67) has
proposed a structure that includes aromatic rings. The cirbon-
hydrogen ratio of asphaltic oils is reported to range trom 0.5 to
0.6 (8, 56, 82).
The mass spctroimetric study of Hood, Clerc, and O'iieal (1.)
gives evidence that the high boiling putrolcum fractions are com-plex
mixtures of rather simple high molecular eightgt molecul.F. Their
work on petroleum fractions of less than -0 carbon atoma indicatrs
that only one polvaromatic nucleus occurs per molecule. Evidence is
inconclusive as to whether this also applies to naphchenic nuclei.
Only one long paraffinic chain is attached to this nucleus; further-
more, there is no appreciable branching in this chain. Other
aliphatic groups attached co the nucleus are short, in fact most
are methyl groups. If this evidence is extended to asphalt a picture
of the lighter components of the bitumen mixture i.n obtained.
C. Fraconacon n by Chromatography
The use of chromatography, or selective adsorption on a
solid adsorbant, provides an experimentally convenient and
reproducible method of fractionating asphalts into components.
Fractions obtained in this manner can be related on a chemical compo-
sition basis since Stc.i,.rt (100) and Busot (1) ha'.' shown that the
infrj-ri: spectra of the sane fraction obtained from Jitf-renc
asphalts arj quite similar although thL relative quanti tcs of that
fraction .iay vary widely.
Kleinschoidt 054) Jdv,.lopPd a chromiatonraphic method using
fuller's earth as an adsorbanc, obtiinLng four fractions called
asphaltenes. wstear-whit: oili, Jark oils, and sjphaltic resins. An
infra-rcd analysis of fractions obtained by Kleinschlmidt'S method
has been reported by Stewart (100) The atEr white oils fraction
v'Ls Sen to be predominatel, saturated aliphatics 'ith some cyclo-
paraffiii and a slight aromatic content. The dark oils arc highly
aromatic in nature rith some aliphatic structure also present. The
exiscence of 0-H, C=0, and C-O groups 6.as also reported in the dark
oils. The asphaltic rEsins ire ec.n more aromatic than the dark
oils and appear to contain more C=0 groups. The asphaltene fraction
shoi both aromatic and allphatic groups. The results also indicate
th, existence of C=0 and C-0 groups but no 0-H or Il-H groups in the
Middleton (70) used alumina chromatography to obtain six
fractions: saturates, mono and dinuclear aromatic oils, polynuclear
aromatic oils, soft resins, hard resins, and asphaltenes. Corbett
and Swarbrick (17) used Porocel to obtain three fractions:
asphaltenes, aromatics, .ind paraftinics + naphthenics. Other
chromatographic methods of separation have been developed by Schwartz
and Brasseaus (1 ), Kask and KIorv (52), Ha'vens and Laniels (3' ),
Kikuchi and Minoro (53), Schvey.r, Chel on. and Brenner i93),
Psalouschchikova (S4) and Zichmzrann (112). T'he asphalEt used in
this research vere tfractionaced by a m. echod de'.'elJped by Sch'tyc'cr
and Chipley (94i. Thc procedure of Sche-yer .and Chiple;, rhich is
a modification of that of Ccrbetc ano S-arbrick, utlllzeJ P.,racet
chromatography to obtain four fractions: asphaLtencs, high aromratics,
lot aromaeics, and parffiinics + naphthcnics. 1he asph-tlrne frac-
tion is Lhat which is insoluble in nonral h:xane; the psraftinic
+ naphthcnic fraction is that hhich is cluLtJd frmn Porocl ic'th normal
heptant; che lo,' aromatic tractlnn is cluted with benzene and the
high aromatic with butanol.
D. Fractionation b; Solvent Extraction
Fractionation by selective solvents has beln -i popular methud
of separating asphalt into components Mctho.ds of separation emploing
this technique have been developed b HIlotberg and Garris '45),
ListEngartern and Sarukhanova (o0), Hubbard and Jcsnficld k-'9l,
Carwin (33), and EisenLuhr and Wirth (25). The asphalts u.se in
this research werc frjctionated b5 a method developed b. lraxler
and Schweyve (103) into asphaLtics, saturates, and cy.clics. The
asphaltics is that fraction precipitated from normal butancl extrac-
tion at 122 oF. using a solvent to sample ratio of 20 to I. The
saturates are the raffinate from an acetone e:ctractiun of the normal
butanol extract using a solvent to sample ratio of 35 to I at -100 F.
'he cyclics are the remaining fraction which is r-rcocrcd from the
3ce one exEract.
Rombrrg, Nsrnth,oand Tra.lcr (86) have analyzed the ira.xlcr-
Sche,.j:r tractions of three asphalcs of i:idcly varied composition
for percentage carbon in aromatic, Cl13 and CH, group. They found
thac the major part of the aromatic structure 'as containLd in the
asphaltic fraction and thi l,-ast .rounc in the saturate fraction.
About three times as much carbon was contained n CH, as in CH3 for
all three fractions of all three asphalts.
Gardn.-r and his couork',rs (32) have developed a method of
fractionating asphalt Einploying thermal diffusion; Lystkhina (62)
has developLd a method of successive coagulation, and Csanyi and
Bassi (l) an electrical potential method. O'Donnell (7?) has
developed a method of separation using high vacuum distillation,
chromatography, thermal diffusion, selective solvents, urea
completing, and chemical reactions.
E. Asphalt Reactions
Asphalt underTgos a number of chemical reactions, omee of
which have been studied but the majority of which have not or, indeed,
even been defined. The most common asphalt reaction is the air
blowing reaction in which asphalt reacts with oxygen in what is
thought to be primarily a combined dehydrogenation and polymerization
reaction. Asphalt also reacts cith other elements of Croup IV-A.
sulphur, sclinium, and tellurium () Sinc.: the reactions .ich air
and sulphur are those that wtr,. studied in this restirch th.-, will
be discussed at length later.
Asphalt has b,-en reaclid -ith hydrogen chloride, boron
triflouride, and silicon hexaflourid, by Ncllensteyn and ilareeu. 7_5),
carbon dioxide by GCrbalinski and Sertienko (31). .nd hydrugn by
Szucs and Ackermann (101) tHilogcnb have been reacted iLth asphalt
by Mariano (od), ;ILllenste,n and Dorle.n (7-), jnd fiircus on (67)
and found to give both addition and substiLttion products. Stevral
process for reacting halogens with asph.ilt h.v.. bejcn patented
(2a, 21, 72.),
Asphalt can be reacted iiLh sulphuric acid to fomr sulphonat.-s
and sulphuric acid esters according t l Harcusson (oi and Guirvitch (33).
Nitration fonns products of the nature of ph.nril nitrouacthanL and
nitrohydrocarbons and oxidation with permanganate forms acids. Thr-
asphaltenc fraction is knovn to form moril.cular complexes with such
inorganic compounds as ferric chloride and mercuric bioToad (') .ind
the cyclics present in asphalL are known co undergo aldehyde condcn-
sation (73). Other reactants that have been used -~L h isphilt art
alkalics, phosphorous, nitric acid, photphoric acid. formaldehyde,
furfural. and metallic salts (2').
F. [he Air BloUwin RL.Ccion
The air blowing reaction consists of many different :.pes
of reactions taking place simultaneously at the intertace of the
liquid asphalt and the air bubble rising through the asphalt. bince
this reaction is of such industrial l rportance, there are many
ref.rcncc-s in the licerature concerning air blowing, but few of them
are concerned with the chemical nature of the reactions taking place.
jLinctic studies of air blowing hi.'e b.:cn undertaken b,.
lHolm ren (-6) and Ariet (t) A loi. energy of .icti.'ation value indi-
cates that the reaction is diffusion controlled, especially, at
lower templrratures. This is further substantiated by Rccorla'a (35)
finding that agitation during air blowing reduces the blowing time
considtrabl,'. Lockuood ( 1) reports that the reaction car be con-
sidered first order except at high temp ratures and low flow rates.
Arite (6) found that at a temperature of around 550 OF. the rite
al[o becoucs a function of the concentration of the asphaltic
The air blowing reaction is thought primarrily to consist of
dehydrogenation and polymeriration reactions long with the formation
of oxygenated h:,drocarbons. A comprehensive study of air blowing
has been presented by Goppel and Knotnerus (34). They report that
the major part of the oxygen reacting with asphalt is found in the
exit gas, mainly as HiO but to a small extent also as CO,, which
agrees with Labout's (57) report that is much as 80 to 90 per cent
of the reacting oxygen is converted to water at high temperatures.
The remaining reacted oxygen is bound in functional groups in the
asphalt, primarily hydroxyl, acid, carbonyl, mnd ester groups. No
ether formation was reported, a finding that contradicts Marcusson
(o4, 65, b6) and Sjchanun (91) but is subscancicted bY the i.ork
of Liiv (59). The proportion of the reacting oxvygr .-hich bLLonmrs
bound in functional rroupa is reported to be a function both of
the composition ot the asphalt and the processing conditions.
Asphalts of loi aromaciLcy can bt dJhydrogcnated to a larger
extent than the more aromatic types, but a larger percunta:e oi
oxygen forms functional groups iir th. aromirtic t.,pe asphalrs.
Ihe formation of escer bonds is of particular importance,
not only because they account for nearly 60 per cent of the o.i.gcen
in functional groups, but also because vstEIr link3g, TiLr., be one
of the more important polymerization reactions oLCiurrinrg in thI air
blowing process. At a temperature of 250 C. Goppel and
Knotnerus (34) report that as many ester bonds as direct carbon-carbon
bonds are formed, although at 350 'C. five times as many carbun-
carbon bonds are formed. The reduced number ut ettcr groups is u-il1
as acid groups at higher temperatures is attributed to the decomposi-
tion of the carboxylic groups. This is verified c:pcrinmentally by
the increasing concentration of water and carbon dlo.ide in the e:(it
gas at high temperatures.
Other reactions that have been proposed by Marcusson (6-) to
take place during the air blowing reaction are the tornation of
asphaltogenous acids and anhydrides of a somewhat cyclic charactLr,
the ox id.ili cn of h'Jri,-carbons to h:,dro>.;l dex ivatives, follo'jEd by
separation tof .ater to form .)n ether, rand the incer molecular loss
of carbon dioX.LJd from an anhvdrLde c .' yield a kcton-.
Anjlysc- of the changes in the various fractions of asphalt
during air bLiaing ha'., b-en riportcd by Eng, Clo-'.:r, and Quon \2b),
Kl'_Lnchimiudc and Snol:k (55), and Gun 36). All reports indicate an
increase. in the conc-rntLation of asphaltrnts, an increase in molecular
weishli, and .)n increase in unsaturation. but aside from thcsc their, is
no concensus on the effect of air blowing on asphaltt compose ion.
S,-rgienKo and hib coworkers (9 ) haj.. found that paraffinE,
c'.clopiraifins, and mo)nocyclic aromatic hvydrocarbons are converted
to conrse'd bic'.clic hydrocarbons, polycondens,-d aromatic hydro-
carbons. and asph.ltenes. Pol'.mriz-ation by,' cross linking is reported
by Mourph: (L7) along U.iih he formation of v.olaLile products by
oxidativc cracking. Gun (37) his shown that the naphthenic rings
are susceptible to dehydr.)genation, but not the side chains. How-
ev,'r, the side chains are susceptible to oxidation according to Eng.
Glovier, and Quon (26) and Zabauin (Lll).
G. Cacalytic Air Blowing
The literature on catalytic air blowing consists almost
entirely of either proposals of various new catalysts for air blowing
and thUir cffect on the blowing tirme or the physical properLies of
chc product, particularly the softening point-penetration curve, or
reviews of the subject of catalytic air blowing with [he emphasis on
the various types of catalysts that have been used and industrial
processing conditions. There is some doubt as to whtchcr the various
additives are actually catalysts since some of them are believed to
combine chemically with the asphalt during air blowing. Alexander
and Shurden (3) report that the chemical stare of ferric chloride is
unchanged in the air blowing process and can bi recovered by extrac-
tion with water, while phosphorus pentoxide seems to become an
integral part of the asphalt molecule and is not water retractable.
Shearon and loiberg (97) have found chat as air blowing proceeds,
discrete particles of P205 disappear, presumably to form a water
Hoiberg (44) has written a rather compreh-nsivr review of
catalytic air blowing, in which he lists the effects of 31 different
catalysts on blowing time and penetration of the asphalt. Howivirr,
the questions regarding the types of reactions that these catalysts
preferentially retard or accelerate to obtain the desired change
in properties remain totally unanswered.
SH The Sulphur Reaction
Theiiii reaction of sulphur with asphalt likewise has been given
little attettnt it is generally thought to be primarily a dehydro-
gInatLion reaction with some polmerization and addition to ,the double
bonds forced by dehydrogenation.
Although it is generally accepted thac sulphur dLhydrogLnares
the .iophalt to form oltiinic double bonas, it is also known that
olcfins art quuIc reactive with sulphur. Brjoks (10) reports that
clefins are nto iound among there reaction products of paratins 'inth
sulphur in liquid phase r,-actionr, probably because- the oliftns
react faster with sulphur than saturated h-droc.jrbonb.. Wsctlak (107)
LCtts C.eidincE of Fricdraann(30) and Armsirong, Lictlc, and uoak (7)
to csho.., chat thL sulphur adds 5lpha to the double_ bond of a l ng
chain hydrocarbon forming a mercaptan intirmnediati which re-acs with
another oli fin to for-m n unsatur.at,.d sulphide. This process could
then be repeated, resulcing in poly mri nation by sulphur linkage.
Horton (4-*) Ijas studio th.- riaccions of alkylaramatics with
sulphur and has found that ihen ulkylaromatics are heated at 200 to
250 'C. '-ich sulphur, dimerization reactions occur. The first reac-
tion h, reportL involv,-s the carbon-carbon linkage oi two aliphatic
side chains with the evolution of hydrogen sulphidc. The second
dimerization reaction is similar except that linkage is by an olef2nic
double bond. The third polynerization reaction described can be
accomplish-.d by four mn.thyl aromatics linking to form a thiophene
tcpe srructurt with aronatics at each of the four carbons of the
chioph.:nc ring. Since asphalt contains many alkylarontitc cype
compounds, it is possible that these types of reactions are at least
partially responsible for the polymerization that is considered to
Pryor (63) report, that sulphur abstracts hYdrogn from
organic compounds at 200 to 300 OC. co give stable products by
aromatization and iinn formation with most of th, sulphur forming
hydrogen sulphide but some being incorporated into crh product.
Ruzicka and M'yer (89) and Ruzicka, Meyer, and Mincazzini (90) have
found that naphthenic rings are dehydrogenated all th- way to aromatic
rings at temperatures abo:.e 200 C Side chains of sulticient length
(four or more carbon atoms) can be closed to form aromatic struc-
tures or thiophene type structures. rhe latter is more conanon in
aliphatic side chains. the rate of reaction of sulphur uith paraf-
finic chains generally increases with the lIngth of the chair and
the amount of branchiness according to Brook.i lu) Higher t.rmpera-
tures (above 600 ')C.) are required for complete oxidation to carbon
I. The Heat of Reaction
The heat of reaction of air blowing has long been of interest
to asphalt technologists because of th. fact that in industrial air
blowing, the removal of the exothermic heat of reaction is an impor-
tant processing problem. While a great deal of speculation has been
made as to the magnitude of the heat of reaction, the only work
published on this subject is one by Smith and Schweyer (93) on an
empirical basis using the change in the ring and ball softening
point as a parameter. Such a determination is valuable for design
calculations but of little value in theoretical considerations.
Research on the heat of reaction of other asphalt processes has not
been attempted on any basis.
If a meaTuremint of the huat of reaction of procLssing
asphalt[ could be obtained, then its rLignitude would lend some insight
inco the nature of the reactions occurring. If the value calculated
f;r the heat of reaction of a proposed reaction from the heats of
formation of products and reactants is of the order of magnitude of
the experimentally measured value, this would be Eurther evidence
that this trpe2 of reaction might indeed be occurrin.. However if
the calculated '.vlu of the heat of reaction deviates greatly from
the cxperimentali.' ocasuied value, then perhaps these reactions are
not as important as the.:l had been proposed.
Furthermore, if the cx-perimentally measured value of the
heat of reaction was different for asphalts of diiference composition,
then it would follow, that different types of reactions take place
when processing different asphalt or that certain reactions play
a more ifpurrant role with different asphalts. The effect of pro-
cesAing conditions such as temperature or extent of completion of
the reaction on chc magnitude of the heat of reaction could be
determined and related to a variation in the types ot reactions
occurring since different reactions would have different heats of
reaction. In this way additional information concerning the reac-
tions occurring at different temperatures and at different stages
in the reaction could be obtained.
Th.- effect oi catalysts on Ech mn gnitud. .*i ih,- li. jt ol
reaction iLS '4jIic Irip.irEunt since : not one but man:' reactions are
occurring during prc,--ssing. If th addition ?t a cui ala,at chil 6. S
rthc ragnitudJ- c iLh: lih-at of rjc iLor," this cculd IcnJ a d tfLnLte
insiht into ii preferrrncial acceler3tion or r jr.Jjt. ion *if
certain c;tpcs of reactions
E.PL-ER II IET rAL
A. D-.tLgn of the Process
The ;.-pcriiientl .ork '.1 c.irrin.d ouL in Lquipm~nt specfil-
call: dLSi16n.lJ and built for th i purpose c.f I. isurLng th'. heat of
r,.acrion of .jphalt air blou.ing, c.aral:'tlc ir blo ing, and sul-
phuri:aci...n. Th- reactor itself 1.a Jd:signLd Lu rLa.t abouc five
gallons ,j asphalt in a batch prrociss. A.dJicional pr..c-ae equip-
rn.n incluJde a drying column to remov'.e ch: moistur, from the
incuoiring .i1 ind .in .ir pr-:h.harer to heat the inlet air to rcac-
tion L .mptriperaturL.
The rLmruv'd of the a.rossol that is forind in th-' exit gases
cf che air blot.ing r_'a:tLion .,a nea-lC sarr, before Lhe txit stream
cuul. b, run through 3 floiieter and1 a g-is analvzLr. Two addi l onal
pieces of Jqui pmTnt '-.er requitrd to soIe this difficult processing
problem-. ith first was an electrostatic precipi..aor, hicih, with
in applied pJcnr-ial of 30,000 volts on the clectrode, would remove
about v9 p, r c.-nt of the aerosol, which would then collect in the
bottom. of the precipitator as an oily product. A srudy of the
chli.mial nature of these oils his been made by Busat (13), wiho
postulated that the aerosol was form-.d by the condensation of th,
more volatile components ot the asphalt, whiih ,.iporiLe Juring the
air blowing. The small amount of aerosol Ith t vas not removed by,
the precipitator, which would still foul the jnal-sis equipment, vus
found to adhLre to glass or claldic surfaces on contact. Conse-
quently a column tilled with ceramic Bc rI saddles subtsequcrt toi
the precipitator would remove all detictable traces of ch- aLrosol.
Two heat exchangers to cool the exit gas st rna, one prLceding the
precipitator and one betwucn it and th- terl addle column, .err
included in the design. An exit gas analysis by-pass line th3t
sends the exit gas and aerosol directly to the sack was adJded to
permit continuous operation of the reactor Lhrough short time
failures in the precipitator. A picture of the equipment is
presented as Figure 1 and a flow diagram of the process is included
as Figure 2.
B. Design of Equipment
The reaction vessel is at stainless steel construction, a
section of schedule 40 pipe, 14 inches in diameter and 20 inches
it height. When charged it is about half full of asphalt, the
IrcIminitng vhl:me being available for separation of entrained
iasIphalt. irom the exit gas. A 3000-watt Chromalox immersion heater
lypire MifiS-230 LU installed in the side of the reactor for the
IIprpaiiie if Iatine the asphalt charge to reaction temperature.
lllA tharmeiiB lll inI the reactor side permits the measurement of the
-- ... ........
rraccion rcmpcrjcurt. A gas disporscr, which is l1,2'-nch stainless
stELl tubing irth 1'32-Mnch duowniard facing hol.s spaccd 1 inch
ap irt, is cold in the baotom cL the reactor to ensure that
complIre iLaxing cakes pldce. A 3 '8-inch sampling outlet i. insctalld-
rnLar the bottom of the riaccor ani a 3!/-inch outlet in the bottom
perriTs draining at the completion of the run.
Th. reactor L ccvir is a flat l .'-inch stainl...s srce l plate,
Ihich is bol ted to rhe reactor. The outlet line for the ,xit
gases i a 2-inch pipe in the top at che reactor. Its largi si.e
runirmides an, entraruinent of liquid asphalt in ch.: EXLe gases. In
this CXii line is placid .a sulphur injection port that pro.,idie the
means ot adding liquid sulphur to Lhe hot asphalt. A Black, SLValls,
and Erys,.n safety hiad iith a 2-inch aluminuin disk in this line
reduces the pos ability of an explosion Gue to an increase in
pressure. The reactor is comipletely insuliatrd with six inches of
Johns-HMn.nville Thernnobestos insulation. A drawling of che reactor
is included in Appendix A.
the air prcheater .as designed to elevate the temperature of
the incoming air to the temrperature of the reactor. The preheater
consists of a steel box containing six compartments. The opening to
each compartment is so placed that the incoming air must go through
a tortuous path from coipartment to compartment before leaving the
pr..he3ter to achieve maximum heat transfer. In each compartment
T Fur e r.' .: r i r,- t A ar r .;
is a 100O-wrtt conical heacinb element. these heat.ing elui-.nts ar,
wired in series to three 220-'.o1 circuits, [w.o htatras in .ach circu t.
io keep the box aircighr the current carrying high crniperature fib,.r-
glass insulated nichromr .ir,_s cnt..r ch b.. through c.rajmic scaled
Cons; chinmocouple glands. The only control on two circuits is oft-en,
but [th oth..r cTirLil coIItains a '.oLigc rtgulacor so chja the .voliEae
of r.ls circuit can be controlled over the onrire rarigie of 0 to
220 volts. "emperacure control of the air entering the reactor is
maincainLd by this volcagc control of th. prLhcater heating elements.
The priheactr and gas line lading to tLhe rc.ctor are also well insu-
lated. A drawing of the preheater is shown in Appendix '..
The drying column .-as dcsiAned to c.-nL in a quantliy oi slli i
geL sufficient to dry the quantity of air that i.ould be used in the
longest run, thereby eliminating th n,.ctssl~ i: of re cr. i.'atLn rh~
silica gel in cth middle of a run. It consists of I 4-inch .Lucte
column 51 Lnches long on an angle iron stand with a couplingr at th,-
top to the 1/2-inch ILne containing air from the cir ompre~sai, r,Juced
to a gauge pressure of 25 pounds per square inch.
Before entering che prcheater the inlet gas first passes
through a Fisher-Porter Precision Bore Flow. rcLr. Th Inle-t gas can
be either the dry air from the silica gel column or nitrogen. 'lhe
air is used as the reactant and the nitrogen is used as an inert tc
mix the asphalt prior to initiating the reaction u:ith air or sulphur,
to mix the asphalt at the completion of the sulphur run, and to obtiun
cooling curves, by which the heat loss of the system can be evaluated.
by meLanb of valv.:s it is possible to si-itch immediately 1rom ai r
operation to nitrogen operation.
The electrostatic precipitator consists of a 4-inch sce.-
pipe WO inches in lnrgth A bakelicE cover is bolted to a flange
at the top of the column. Through this cover and a high voltage
ceramic bushing is extended the high voltage electrode, a
l/'-inch stuel rod. The aerosol enters the precipitator at the
bottom through a baffle arrangemitrit that ensures that the aeoisul
is well distributed in the precipLtator. The loetr ex:trcmity of
the clecctodC is secured to 3 gIL&5 insulator attached to the center
of the baffle. The precipitated 3crusol collects as an oil in the
bottom of thC precipitator below the inlut baffle and can be removed
by opening a valve at the bottom.
Th,_ voltage source is a Carpco Model RI high voltage rectifier,
which supplies a voltage that can be varied from 0 co 40,000 vales
with a current output of 12 millLamperes on dead short circuit to
The packed column containing the Berl saddles to remove the
last of the aerosol was constructed from two sections of pipe joined
by a bell reducer. The lower section is a 3-inch steel pipe and the
upper section is a 2-inch steel pipe. The total length of the
column is 73 inches. A drain in the bottom of the column was included
to perdi t drainage of the collected oils but this proved useless since
the oil adhered to the surface of the Berl saddles and would not flow
to the bottom. Consequently it was necessary to empty the column
and clean the packing about every three or four runs.
The heat exchangers are the concentric tube L-pe -itLh
cooling i.atcr in counter current luow on the shell side. The firr,
ihich precedes the precipitator, consists of a 1 i1-inch steel pipe
inside a 3-inch steel pipe uO inch-i in length an'j the a c.ndj, which
follo,.s th, precipicacor, Is a 3 2 -inch steel pipe instd a 2--inch
steel pipe ,S inches in length. Since .-ater cc.jnensed in ite i irst
exhchangrr anld a limitci amount of aer.se-ul prcipicated in both
exchangeLrs, \'al,.es r.'t. instal ld in th.: b,'tto.)m ends of the heat
exchangers. The entire apparatus is mounted on 1 1 '.' ;: 12 .
1/-inch antgl iron framcw ork. Drbra'ings of thi heat cxchaners as
well as the electrco.tatic precipitator drying colorin, ndnJ surfidc
column, are included in Appendii A
C. Exprr-i mental Deig.n for ir Llov-line
Ihe air blowing e:.-pcriments .were statistic ll lJeaigned
to measure the effects of three factors on th, hcat of reaction. The
first factor ,adS the type ut asphalt. For thih, three different
asphalts were selected: an EIdt ITexas aphJltic rcaiduum, a Gulf
Coast naphthenic residuum, and a West Texas-flew Me:.ico riesiduum.
They werc chosen because thc', represent ditfercnt composition ty,'pe
and the effect of the type of asphalt on the heat of reaction, if
significant, could be expecteJ to appear in the results of reacting
these asphalts. A list of th, properties of those three isphalts is
presented in Table 1.
C.s1 TiLxaj Gulf CouSE :5st Tcxis
Asphil ic tlaphthiL:n c l i' HMLXI-C
Rediduum Rijlduum RI:siduum
Idinc Lfic ai n
Specific GravLcy, OOoO 'TF.
I nL.matiL: L.' SC 5 L T.,'
Scok.-s, 210 oF.
PErc traction, 77 OF.
Ring and B-ll Soictning
Sulphur Content, '
Oxygrn ConciEn, 1
Component Analysis (a)
Comrponent Anilysis (b)
High Araomaics, '
Low Arumatics, 3;
Part. + Naphth., %
too soft mla
(a) Uetennin-ed by method of rraxler and Schwcyer (103).
(b) Determined by method of Schwcycr and Chipley (94).
ThL second factor co be studi.-d .wa the e':-tent :f the ct m-
pletion of the air blowing process; chat i1, to detErmin i t ch h. at
of reaction changed significantly a th. air blowing proced>dl As
an indication of the .-.trllnt of i.ompl clion, the rinL and tball softening
p.,int Was selIected. The ring and ball StofLtninf point is ui.jasur-d
b:, a method described by the Am..rican Socitt7 for Testing tlacerL3al
Test D 3j-2t. Tht softening poinc 1i 3 measure of th- conslsr5Cncy
of the asphalt and is known to increase appro:imiatel lin.early Lich
time as air blcuing proceeds. Fror Table 1 i I can be se, n thiL th.
soft 'nini point of S-b2-3 is 70 oF., consid rabl b lo-. thost of
S-o2-2 and S-b3-3, which are 9b an. 100 oF., rspctively. Th-
levels of softening point thea wLre considered w.-re 100 co 130 oF.,
130 co 130 oF., and 10S to 250 0F. for all threL asphaljc. In adji-
tion the range 70 co 100 oF. was considered for asphalt S---J.
The third factor to be studtJid %.as reaction trLanpiturce.
Since the e:.pcrimental method requires appro:xir ae adlabatL i op,.ra-
tion with the only loss of heat due to normal convection h-at losses
trom the outside of the insul3aton, isothtrnMrl operation was not
feasible. Therefore two ranges of tCtperatulre w.re selectEd as the
temperature factors. The- first range was from 4tO to 475 'F., which
is the lower end of the termpwrature range at w-hich asphalt is usually
processed and the second was from 535 to 550 F., which i-a the upper
end of the range.
For this type of operation a factorial .xpErimwcnLal design
will reveal the most statistical evidence. iwich a minimum of L.,peri-
mental data. A requirement of Lhe factorial design is that all levels
ot death factor must be the aunt for all other factor,; that la, th,
tempLrature and SuiLening point range- must be identical for all
three aaphalts. Ihis condition is satisfied if the softening point
range 70 to 100 OF. fur asphalt S-t2-3, is not colnsAierd inI the
ana1ily i of 3arianc,. In this factorial dJsign th-rT are three
levels of asphalt, three levels of softening point, and two lev'.ls
of tcmpcrature, hence it is dtstgnated a 3 :.i 3 2 factorial
Thi anal',sis of variance for the factorial reveal the signifi-
cance of both simple effect and interactions. A simple cffect is
the difference in response over different level. of the siej factor,
uhile an interaction of t'o or more factors is the differencL in the
simplee ciffects of thus factors. If theL e effects differ by more than
can be attributed to chance at the 95 per cent confidencL limtl, the
Jiffc rune is said to be significance. If the difference is greater
than can be attributed to change at the 99 per cent confLdence limit,
the difference is ,aid to be highly significant.
D. Operating Procedure for Air Blowing
A weighed quantity of asphalt (43 to 50 pounds) was charged
to the reactor and the charge was heated until it had reached a
temperature abouL 30 degreci higher than the desired temperatLre
range. Nitrogen was bubbled through the reactor at a low flow rate
to nmil the asphalt, thereby ensuring chat the asphalt temperature
was uniform throughout the reactor. The heater was disconnected and
temperature readings of both the incorning nirrog-n and thv .asphalt
charge -.r,- made c.very, flij minutes iith the nitrogen c--mperature
being regulate d to the same temperature as the a phjlt. lwh>n the
temperature of the asphalt had dropped to the loter limit of the
temperature range (460 or 535 F.), the incoming nitrogen .as rpiac,.
by air, whLt-h ruac~Lcd 'ith the asphalt. The eiXt gas Wen.t through
the h a-t exchangers, electroscatic precipitator, and the pa.-k:d column
to a Fisher-Porter flowmenler, .herLe th.- floow rate of thL aerosol
freer g3as as iImasured. The gas stream was chen split -ith the majuriL.
going to thi stack and the remainder being passed through a Small
silica gel column, where the moisture as well as remaining traces oi
aerosol were removed. The gas then entered a Ecckr-n n Model C oxygen
analyzer to determine its ox,'ygn content.
As the reaction proceeded the asphalt temperature rose because
of the liberation of the exorchunri heat of re,: tLon. During this
period readings were made every five minutes of asphalEt temperature,
inlet flowmeter position, incoming air tempe-raturi and pressure, the
pressure on the system, outlet flo-meter position, e:-.it gas o:',yger
content, and ambient air temperature. Ai soon as the temperature of
the asphalt approached the upper limit of ith temperature range
(475 or 550 OF.), the air was replaced by nitrogen, causing the
temperature of the asphalt to drop again due to normal heat losses or
the system. When the asphalt temperature again approached the lower
lirrit of the te-pericure range, air wao reintroduLced, the rtecJton
has ruinitcated, jnd the temiptrature of the charC. again ros'. This
procedure .as repeated until the r-actci.n had slo ,!ed to the point at
which h the: e:.:otherr.ic hear of r ca3tion Ta s no lon-er suffici..nt to
maintain the asphalt at reaction ti.mperaturc. A ctimr-tetmp-er tuilre
profile from j t .'pical run is illustrated in Figure 3. Each point
on the graph is jn exiperimtcnac value of the asphalt temperature.
During the latter part of the rca tiLn jt thL lower Ltmpers-
ture range, the race of reaction 'as not 5uiffiicLcntl rjpid to keep
th2 temperature above the lo-er limit th th tcrmprature range. In
this cias it w5is nAcssar, to reheaC the asphalt t- the upper limit
and obtain data as the system cooled '.hlle reacting until the
softening point of 250 oF. 6as reaJchd.
Samrplcs 'JLre t.li.n approximately ever' hour during the reac-
tion and were rmjsured for ring and ball softening point. Selected
saraplcs iL re nal,.'::d for o;-':gcn concent to decenrine the percenLage
of rccticng oxygen actually cheraically bound in the asphalt. Thcse
values are reported in Appendix 1.
rhc amount of heat liberated during eich incremental rise
could be calculated as .,wll as the total amount of oxgenn consumed.
In this 'ay the heat of reaction per mole of oxygen was calculated
for each increment. A total of cveLve runs was required for this
part of the research, two duplicate runs of each of the three
asphalts over the two temlprature ranges. A table of all the ezperi-
mcnLal dJta is given in Appendix 1.
E. E.'primenral Design for Catalytic Air Bloting
To determine the effect of cat3lysts on the h-at of reacttl-n
ut air bllo.irn asphalt, ri.o of the most common catalysts, alumireum
chloride and phosphorus pentc'-.ide, uAro sleccctd. A third catalyst,
ferric chloride. was also sclicted but experimental difficulties
forteil the discontinuance .o t.ork with chis catalt.st.
Since the cype of asphalt, t ir.elIatuie, and extent of com-
pletion of the reaction utwre not ficcors in this part of the research,
a dLtf..r..nt statistical design was required. The most meaningful
results iCii the nunimum of experimental data could be obtained by
applyLnp a method developed by Dunn.ctt (23). The objective of
Dunnctt's procedure is to compar- sev.'ral EreatmenEs with a standard.
For chis study the standard L.as taken to be the heat of reaction of
air blun.ing asphalt 3-62-3 over the softening point rangc of
70 to 130 OF. Since two duplicate run3 of air blowing S-b2-3 at
the temperature ranges of 4oO to "75 OF. and 535 to 550 OF. were
complr.ted in the firsr part of this research, and, as can be seen
in Table 8, there tas little effect of tenperacure on asphalt
S-o'-3, four replicate values of a standard had already been obtained.
EB air blou.ing asphalt S-b2-3 over th, temperature range 460 to
550 OF. wirh aluminum chloride anil phosphorus pertoxide with
duplicate runs for each catalyst, the five degrees of freedom
required for Durnert's test are obtained.
The -xpcerimr.ntal proc..dur for catalytic air biornin& i
identical to that for ordin-rry air blowing Cxc-pt that 2 p.-r cenc
catalyst '-a added t.J he asphalt and th3a th- reaction was
IniLLi. ted at a timperajure of 4bO "F. and ailloc.Ld to .1I.'JtiL thc
temperature to 55U oF.. at which point a sottcnirit pointri af ppro::-
imately, 130 OF. had bi.n obtained and the reaction .35s ciTr-inacd.
Th'e eliiiination of temperature as j iftn; r and C-.rnsqu ntl1
the need for altcrnatc nitrogen and air operation, as W'el as th.
terrinailon f the run at a softening point of 130 F. instead of
250 "F., considerably decre'asbd the. tmir rLquird to c-oruplctc an
c.-pcriment al run.
F. E:ipcrimr.ntal estzin for SulphurLzation
Because the reaction of asphalt itch sulphur Ls lnduthcrrIc
rather than exotherric, as is the case in air blci-ing, and because
sulphur reacts as a liquid r-th.r than a gas, a ditiercnL cxperi-
mental procedure was required. The only factor to be ccnsidErEd in
the sulphur reaction was the repe of asphalc. Ltch che hena of
reaction of the three asphalts, the E5SL TeXS., thL Gult Coast
naphthenic, and the West Texas-clew le.
an '-.xperimcntal design uLilizing the F-itL tas appropriate.
In the procedure adopted, the: ruictor was charged and he-ated
to about 575 OF. as in air blowing. Ten parts per million of Dow
Corning Fluid dissolved in 30 milliliters of kerose-re was added co
eliminate the foaming that was found to occur shcn the asphalt and
sulphur rea.ted. Nit rogrn war bubbled through the asph lt .t a low
flo' rate to ensure uniform, t mriperaturr. [he heat was cut off, and
when ch. temperature of the asphalt dropped to about 550 "F., che
nitrogen was shut off and a weighed quantity: of liquid sulphur
(about 1200 Frams) was added in approximately 200 gram increments.
The ensuing reaction was very rapid with hydrogen sulphide giv'.n off
at a high rate. After th-, aerosol mist in th. exit gas, which
appears to a lesser e::tent than in air blowing, had been removed,
the floI rate was measured by the e-:.it gas flowneter. In a few
rrminutL the reaction would subside and an additional increment of
sulphur was nJddd. This w.is repeated until ill of the weighed
quantity of sulphur hid been added.
When no further evolucion of hydrogen sulphide could be
detected (about 15 minutes after the addition of the last increment
of sulphur), nitrogen was readmitted to the asphalt to keep the
system ''ell mi:ed and at uniform temperature throughout. Tempera-
ture readings of the ,system wr.-: recorded every five minutes and
the temperature at this point was usually between ,70 and 450 0F.
After 10 to 15 minutes of operation with nitrogen bubbling through
the asphalt, it was assumed that all of the sulphur had been reacted.
The heat consumed during this entire operation could be calculated,
hence the heat of reaction per mole of sulphur consumed could be
determined. The asphalt was reheated to 550 oF. and the procedure
repeated with another weighed quantity of sulphur. In this manner
fifteen values of the heat of reaction were obtained from six runs.
A. The heac Balance
The quantity of hat liberated over in incremental reaLclon
ran.e ,ab dc tt Lmin-d by making a heat balance .n tile sstel.
Input Output + Ceneracion = Accurmjlation
H- H L + Q = As + Ar + A i)
Q = A + A + A 1H H ) + L (2
as r i a
Ha = nthalpy of cnte.;rin air, BEF
Ho = Enthalpy io exit gases, BTU
L Heat Losses, BTLI
Q = Heat generated by the reaction, BTU
Aa = Accumulation of .ncrgy in the asphalt, BTU
A = ALcumulaLion of energy in [he readLing .esssel,
BTUAcumul o nrg the nsl.
A. = Accumulation of energy in the insulatin, BIU
i. F.aluation of Input and ClUr puL T-rms,
ihL input ann output terms of the h-.at balanc,- .re i'.jaluacedl
in thh-- following manner.
h = F C (T ) AE (3)
,1 a a J
'-h r c
F = tias. flo. rate of air, lb. nun.
Ca = He. Ic p 3 iC 1 :. 1 f Lir, E[I. I b. 'O'F.
Ia = Tcmp.-raJt're of incomiang 3 ir, F.
T = Arbit rary Jacum ceimpcrature O.
t = TiML, miTn.
Ho F FCI(T T ) T d
F = Mt:.as floi. ratc or e:4it gas, lb. rirni.
C = He'i capacity of exit gas. BTU lb. OF.
T = Temperiature of exit gas, OF.
Since thi. .x:act composition of the exit gas was unknov.n, the
''alue of th.. h.-ac capacity of the 'exit gas scream could not bc
deLtrrdlned. The asuiaiption was inmde that
F C = FC (5)
SLnce the incoming air is 50 per cent nitrogen, which is inert,
and no more than 30 per c,:nt of the oxygen reaccs, this assumption
could not bc in error by more than 7 per cent. It was also
assumed that thcrmnal equilibriu-, was approached betcL.en the air and
asphalt during [h,- reacting period so that the temperature of the
..xi gas '-as ch,. sar, as th- ttmpcriraure :- the -sphal that is.
1 = T ) In thI.roughl iii:d:...l systc m rthi assumr pctLon is .1 va.-il .
ia- "fo = FaCa(Ta as'I (
T = Tinp.raLure of thc asphalt, UF.
BEy mTaintaining the inlet air ermperature close to Lh. ct._mprarure
of chL asphalt, this CEnr can be mffAii small, juLscit)in th.e assuirin-
C. Evaluation of Actumulacian ermis
The accumulation cen is air eUvaluaiteC by th_ foli'-iLng
A = M C LT, T ) (?)
a h,.S ,
M = Mass of the asphalt charge, lb.
C = Hea3 capacity of asphalt, BTIU lb F .
T TIiriperacure of thi asphalt at time c, F.
T Tremperature of the asphalt at time t + 8 c 'F
Ar t'Crlr ) -
Mr = Mass of the reactor vessel, lb.
C = Heat capacity of the reactor ve-ssel,
BTU b. F.
BTU lb." F.'
In -riting :rquatton (8) it was a-i;umI d that t th th..rrial condue-
tl r'- of thL sC3 winless sL-l reactor an the asphalt film coeffi-
cient w r, infinite, c.auing thc tcmpEr3turc ui the entire vessel to
be the sami is that of the reacting asphalt. This assumrption might
seim to be a rather poor one because th-. film transfer coe ficirntc is
known to be Iow. However, if it wcre as low as 10 BTU p.r thliur-
d,.-grc F.-squar foort), the magnitudc of its resistance would bi
small fr in o,.rll l h'at transfer co.ifcitent of about 0.2, whichh
was calculatt.J for the insul:tcd reactor, so the assumption ..ould
The iirst ter i in equation (9) is the accumulation of energy
Ln the Insulation on thL iides of the reactor and the SLcond teni is
the acciijuul.ation n the cop and bottom insulation.
A, = 2Trc h I r(r) T(r) r dr
+ /2r R i [T T(xl dx
C = Heat capacity of the tnsultacion, BTU lb. F.
h = Height of thL reactor, Ln.
r = Radius, in.
x = Vertical distance from top or bottom of reactor, in.
R = Radius of reactor, in.
R. = Outer radius of insulation, ,ln.
P = Don ity of tie insulation, lb. in."-
T1(r) = Temiperaicrt of che insulation as 3 funciLLn .f
radius at iiiic uF.
Ti(r) = Teiper3turc of the insulation as a function of
radius at time t + At, F.
T (:-.) = Tempraturr of ch, inulatic.n on rht top and
bottom ends ot che r--cctor J a function of
position ac tire t, oF.
T.,(:) = fcrpcrature uf the end insulacior. as a function
of position at tiiie t + Lt, t' .
D. Solution of U.nstiad, Statu Heat rrii:f r Probl~-.r
Considering the first integral in equcaion ('), it .as ntcei-
sary to cvaluatson T (r) 3nd f (r) as functions of radius b>forc th.
integration could be carried out. he.-e vari3bl s can be relaEcd by
setting up an unsteady state heat balance, Ihich yi-=lds a second
order partial differential equation, t-hich is gi'.un in cylindrical
Cip a T T 1 T I 1 if T T f"
-- +- + (10 )
k at ar7 r dr r- 3G' (
k = Th.nrmal conductivity' BTU min.-l .1 in
r = The diStance trorT, the a>is ot th, .:lin,,cL
co 3ny point on the radius, in.
9 = he angle betr .cn r ano the x-axis, radians
= ThL. distance measured along the axis uf the
cylind. r, in.
If the assuJnptiun is ruade that heat transfer takes place inli in
the r3daul direction, that is, f = 0 and = u, th. LequIation (I0)
r -- r cr -- 1
dr-' r k r. Il)
The boundar-. conditLons for equation (LL) L.rr: taken as
Boundary Condition 1: Ac r h ,, T 1(c), .here fit) is
obtained e;..perimentally b, the temperature readings of the asphalt
at il,. r unutc time intervals. A plot of = f(t) for Pun R-o3-b
is shuun in Figure 3.
Boundary Condition 2: At r RF T (rt), where g(c) is
obtained e..perira.intally, by a knoledge- of the ouvr-all heat transfer
coctficienc between the outside of the insulation and the ambient
air. The temperature at P.R can be stated as an explicit function
of reactor temptraturL and ambi:-nt air temperature, both of khich
are functions of time. This calculation is shown as Appendix B.
For the initial coni cion, steady stat.' 15 assumed at
rice = 0. The solution ro iquJLiCon (ll) with -- = Is
the sAultion of the ordinrar difi. fernt Ll tqulrtion
r + J = (I12
iich the bound:iry conditions at r R,, T Tp nid it
r = T= IR,
Initial Condicton: A t = 0
I T T In [ r ,
In [P. 'R,
A finite di fierenc.e jppro...iif, ti n w a ua.sl- tc~ uol.'e
equation (11) with th. preceding boundary and initi l condiCLons.
The finite difference equation i g '.'.-n by
(I + 1/m)I Tr ,n (:1 lIm 2) n Tm- .n
in = r/ Ar
Increments of I inch for Lr and 5 mirrnndts for Le satisf:'
the stability criterion for finit, Jdiffer.nce equations (6') a id
were selected for use in making the calculations. The. calculltions
for stability are presented in App.ndix C.
An B1 709 computer Wass employed .to sole'. the boundary and
initial conditions anJ to make the calLulacioni in the finltA
dtifercnce solucton. [he coripuccr program, lrittcn in the FORFPAN
language, is included as Appendix D. After Tlir) and T1"r) had
be.n obLained as functions of r.lius, the Integration was carried
out numerical ".
A similar procedure could hjve been followed for [th Linsu-
latro n n the top and bottom ends of the reactor but this u.ould have
tii'.olved a conSLderable amiunt of additLonal Calculations. AS an
approximation, the, accurmul a3on of energy in the enrd insulation '.as
assumed to be proportional Lo Lhe accumulation in the side insulation
.irh the ratio cof heat transfer areas taken as the const nt of
proper t ional t,'.
E. Evaluation of Heat Losses
The, rate of heat loss from the system uas evaluated from the
data taken chile nitrogen rather than air was blown through the
reactor. In this case no reaction cakes place and so no heat of
reaction is generated. Nine values of the heat loss were calculated
for operation o'.'Lr the c mpcrature range 460 to 475 OF. and 16 values
for operation over the temperature range 535 to 550 OF. The average
of these valuess was found to b 10.61 BeU per hour and 11.64 BTU
per hour for the temperature ranges *bO to 475 OF. and 535 to 550 F.,
respectively. The heat loss term is then simply calculated from
L L' Lt (15)
L = Calibract d raci: .f h-at loss, BTUi rin.
The value of Q is then obcained siipl, by, suina=n all tcris
on the right-hand idie of equation (2). A chick as t tthI corrLct-
ness of chis procedure W3 Jd..'isd by c,-nncting a '.oirmeter and
an atmrrietcr across Lhe line EL tch reactor headr and m,.j:.urinn the
energy input of tLh h-ater oy'r an Incr.-mir ntal p.:rt..d similar to
one th3at toull be ,:pcricnc,:d in thie air blou~ng process .. Thh i .as
rcp-eaCtt three tines and the results were crTmpi-ar, .lth .'jlues
obtained by the calculation procedure just described. ..-gr.,-ti.nt
within 4 per cent wJa obLtinio. This calculation, alone Lith h1le
loss calibration and fllowmn:t-r and o.a,'gcn anal er calibratioLns can
be found in Appendi.: E.
F. Calculation of Oxy,;-n C'nsumerd
The total number of noles of ox.-:gn consumed during an
incremental reaction period can be determined by numriiriL.lly integration
S=1 (0 209 F, YoFol) d t)
G = Oxygen reacted, lb. moles
Yo Exit gas oxygen concentration, rniss fraction
The heat of reaction is then obtained simply by dividing
the ncgativ.,e ,o the heat generated by the number or moles of
SH = Heat of reaccion,kcal. gm. mole-1
The he. of reaction of the sulphurization of asphalt is
obtained by dividing thi negative or the heat generated by the
number of mols of sulphur reacted.
A sample calculation of one particular heat of reaction
value for the air blowing process and one heat of reaction value
for the sulphurization process is included in Appendix G.
rE SU L S
ni. Air Elowing
The individual results obtained for the he-; of reaction
of air blowing asphalt are reported in [ables 2 and 3. All v.alJEs
3re reported in kilocalortni prr gram molc of o....gen consumed
HEAT OF PEACfrONI OF AIR ELUWING ASPHALT, TEM1PEFR.UR "o00 "75 OF.
Asphalt Heat jf Rcaction, kcal. ':7. molt
East Tcxas Gulf Co.ist lest Tx3as
AsphaltLC Na'-phthitnic eili Me.ico
Residulum Residuum Residuum
S-62-2 S-o2-3 S-03-3
Softening Point -
70 100 OF. -o0.9
Softening Point -73.2 -o4.5 -08.5
100 130 OF. -71.2 -o2.7 -b7.6
Softening Point -71.0 -05.9 -71.3
130 180 0F. -71.5 -bl.4 -07.3
Softening Point -74.0 -57.0 -o9.3
180 250 OF. -71.9 -65.5 -72.6
HEAI UF REACTION OF AIR ELOWINJC ASPHALT, TEFMPERAlTRE 535 550 F.
Asphalt Heat of hRaction. kcal./gn. mole
East T-;as Gulf Coast West Texas
Asphialic Naphthenic New Mexico
Residuum Rcsiduum Rcsiduur
S-b2-2 5-62-3 S-b3-3
Sottening Point -ol.1
70 100 'F. -il.3
SOfLcning Point -67.8 -60.2 -67.4
100 130 OF. -ob.4 -bO.1 -b8.1
So[frning P)linc -t4.6 -60.O -69.,
130 160 F. -b5.8 -b3.3 -bb.7
Sofcning Point -2.9 -b .o3.7 -65.5
ilO 250 oF. -6b.b -60.2 -b7.6
4 facturial analysis of variance was calculated for these
data, excluding the softening point range 70 to 100 UF. The
statistical data are reported in Appendix H. The analysis of
variance of the air blowing reaction is presented as Table 4.
ANALYSIS OF 'JARIAN;CE OF ASPHALT AIR BLOIVINC RESULTS
Source Degrees of Sum of Mean F
Freeuom Squares Squarc
BLocks 1 u.ulO 0 u100 0.002
Asphalt, A 2 37 .511 169 ;5-58 3o. 703 *-
Softening Point, B 2 O.OLlb 0.0058 0.001
TemperatuTe, C 1 83.4117 b3.411 t..134 *
AB Interaction L2.0o-1 3.1554 0.0S4
AC Interaction 2 52.5.07 2o.2704 5.081 *
BC Interaction 2 1.3033 0 bSoL 0.136
ABC Interaction 4 17.5593 3898 0.3'9
Error 17 87.890uJ 5. 1700
Total 35 b34.5000
The single asterisk on the value of F in the analysis of
variance means that the F is significant at the 95 per cent confi-
dence limit and the double asterisk shows significance at the
99 per cent confidence limit.
The most apparent result of the analysis of *.'arinnce is the
highly significant F due to the effect of asphalt showing thac the
heat of reaction of air blowing asphalt diff.-rs for asphalts of
different composition, a difference that is signifiLent at the
99 per cent confidence limit. The signtlicant F reported for the
effect of temperature means that a change of temperature of the air
blowing process modifies the heat of reaction. The fact that the
F for the effect of softening point is not significant shows that
the heat of rtaCLicrn do:s n Lt aigniiicant l change as the air
blowin; reaction proceeds.
The significant F rLtporttd for AC or asphalt-tr mperaturLe
intc.r.ir ttI n mcans chat th.. Jiif.r. nce in heat of reaction due co
rcnp.tratur.- diffurLncc is not the sanm for all asphalts or, con-
v,. rcly, it means cth t the di firencc in th,. heat of reaction of
di ffer._nt asphalt is not the samin. at difl t rent temperatures.
The oth.r incerrctions, asphalt-so.itning point, tumperaturi-
softening poiut, and asph ilt-t.Irmp,.rature-softLning point, are noo
sign fi c icn ,
E. L.i]al'y i .Ai r Blowing
The inJiviJual results obta.in'd for the heat of reaction of
catalytic air blc.'ing chL u,,l Coast naphthcnic residulum, S-2-3,
tLLh 2 per cenL catalyst are reported in Table 5 along with a
standard, the heaL of reaction of air blowing S-62-3 from its ini-
tial softening point of 70 uF. to a softening point of 130 OF. Again
th-. values are reported in kilocalories per gram mole of oxygen
By means of an analysis of variance of these data, which
is reported in Appendix 1i, an error mean square of 2.89 is obtained.
Dunnett's (23) procedure is applied and a Dunnett difference is
calculated at both the 95 and 99 confidence limits by equation (18),
xhich is recommended by Steel and Torrie (99) for use with an
unequal number of replications.
HEAT OF REACTION OF CATALYTIC AIR BLWIN;;G ASPHALT
Hesa of Rcdition, I:Ll1cal1.rLC's grTm MulI.
No CIt Il ysL Aluminum Pho*phorus
Chloride Prenc,:. ide
-c4.5 -71.0 -b3.3
-61 8 -71.1 -LO.
Mean -612-) -tI .b
= Dunn,-tC's di f ffe r, i:.
T Dunnett's t
s = Error ifan square
p = Replicarton s wirh standard
L = R.-plicati._ns tith corp tiison
The 'alui of Dunnett'z t tor fiv.'e dcpgre..s of freedomri is
2.4. at th, i p5 per cEnt confid..nL.- limit and 4.21 at the 99 pLr cent
-ontid. ncci limit, whichh r. sulis in Dunn..tt di f..renccs of 3.5 and
6.0 .3[ the 93 and 9Y per LL.nc L(.-'ni ide nce [v.'1ls, trvpLLLively. From
Tabl-- 5 Ic can be seen that th. m-ans -of ChL standard, straighL all
bloutrig iith no catalyst, and the comparison '..th phosphorus pentcxide
do not diifer by more than 3.5, and so their difference is not sig-
nificant and it is concluded that the use of phosphorus pcnto:iLdc
as a catalyst has no significant effuct on the magnitude of the heat
of reaction of air blowing asphalt. However the mean of the standard
and the mean of the conpartson of catalytic air blowing with aluminum
chloride differ by more than 6.0, hence the difference is highly
significant and it is conrLudrd that the use of aluminum chloride
has a highly significant effurt on the magnitude of the heat of
The individual results obtained for th, heat of reaction in
the asphalt sulphurization process are reported in Table o. All
values are reported in kilocalories per grim mrole of sulphur reacting.
HEAi OF REACTION OF SULPHURIZING ASPHALT
Heat of Reaction, kilocalorits'gr.rm m'l,
East Te:;.as Gulf Coast West T :a
Asphalcic NiphrliLhen. .liw M .xico
Residuum R--sLduumn Ri-Iduum
S-b2-3 3-62-3 S-b3-3
+11.2 +8.3 +6.b
+14.5 +6.6 +b.-
+12.0 -7.5 +6.3
+11.1 +6.0 +3.3
Mean +12.2 +8.2 +.-.0
An analysis of variance of these data, the statiatLcal data
for which appear in Appendix H, is reported as Table 7.
ANALYSIS OF VARIAI;CE OF ASPHALT SULPIHUPRILTIOII RESULTS
Source Degrees oft um of Mean F
Freedom Squ3res Square
Among 2 -1.702 20.851 13.6b A
Within 12 18.365 1.530
Total I 60.ub7
The highly aigniticanL F obiainid indicates ihat the hrat
of reaction of sulphuriLing asphair is significantly dltferint with
different Lype,' of asphaLS at the 99 per cent confidence limit.
ANALYSIS OF RESULTS
A. Air Blowing
The highly significant difftrtnct in the value of the heat ..i
reaction for different asphalts demonstrates chat Jdiifernt types of
reactions do indeed play a more Iripartant role in cLrtjin ct'p _of
asphalt. The lack of significanciE n th. titect of 'aoisorniag poin
seems to indicate that the samn types of reactions occur throughout
the blowing period of a given asphalt. although this cannot bc
stated conclusively, since compensating changes could occur in ihich
the heat of reaction per mole or oxygen Lonsdumad would n.t change.
The significant difference in the hear of reaction for operation at
different temperature levels implies that different reactions do
have a greater importance at different reaction temperjtures. if the
effect of change in heat capacity rmy be ruled out, which is not an
unreasonable assumption over the moderate temperature range considered.
The significant asphalt-temperature interaction indicates rhat the
response to change in temperature is not the same for all asphalts.
This becomes readily apparent from an analysis of Table 6, which is
a condensation of the air blowing results.
fri LE 8
SL lIARY OF AIR BLCOWIIIG FRSULTS
Asphalc Heat of Rc.i Lion, kLil./gm. mole
TeioIpc ru re TerTmpr:ra ture
o00 475 'F. 535 550 "F.
East ITxis Residuum, S-o2-2 -72.1 -65.7
Gult Coast iNphthcntc, S-b2-3 -o2.d -bl.2
WEst rTxas-Nce. Mexico, S-o3-3 -t.9.5 -by.O
Examination of Tabie 6 reveals that the te-ripuratur:e etfict
is c'periencc-d almost enLirely by one asphalt alone S-62-3,
accounting for th, signifiLant asphalt-temperature interaction.
rhe heats of re.ictiun for the various type' of reactions chat
have been postulated to take place during air blowing are presented
in Table 9. ThesL values wLre determined by applying the general
principle that the heat of formation of a chemical bond is a constant
independent of the remaining structure of the molecule. Pauling (78)
has sho.n that this assumption agrees quite well with experimental
values. The calculation for each of these reactions is shown in
Appendix F. The values are reported in kilocalories per gram mole
of oxygen reacting with water as a gaseous product.
CALCULATED I-.AT OF REA.CTIO U VALUES FOR PROFPSED AIR. BED)UING RE.A'TIlls
Propuoed Reaction Hejt of R.-a.-cton,
kcil./-T r moat
DIhydrogenation of single bond co form doubl, band -.5.2
Aromarization of naphthenLc ring to aromatic ring -..
Direct carbon-carbon bond fonnation -'<3.9
Formation of ozx.gtn containing functional groups
Ketone on side chair -9. -
KcEonc vi h naphthLnic ring rupture -L5 2
Carboxylii acid -1-)2 7
Hydroxyl group -' u
Decarboxylation -. (- ', X. '
A correspondence betceon the :zalculatcd hEac of rccltion
values shown in Table 9 and th, expL-rimental v.allus' shown in T.ible 8
does not in iLself justify the proposal nf a mechanism f.r th.-j ir
blowing reaction because the combination of various r.acElois o.ccurrLng
in different proportions could result in almost an' over-all heLa of
reaction value. However, such a correspondenf,-c can provide corroborat-
ing evidence that a mechanism proposed on the basis of additional
evidence does indeed occur. Conversely, the lack of a corrcspondtncE
in an ..:,~ rimLntal h..a.i of r,- ctr i'n '.,alu I .ir.I the C.~ alcul3 ed .I lul
of a propuatd r action mchani sm proadJL I,- Jtdin.. that the propoi.d
muLchani srai ma. not bL as airport anc is hi beLnr pcstuli ed.
Ic is s -.in that .-t,. r ifoniition and ct rbo -i boi- i on link.ge
havi. n arl., the sart. hear of reaction, hich ni-ns that no inf',crma-
titoi will be rccealed concerniLn .-hicn at the-c t%.o postulated
polym eriZ3t Lon linkl gc may: be' fi'.ored un-der certain conJ it ias. It
is interesting to note that th- onl,, proposed reaction with a less
negici .rc .'3alu'e t the heat of re-ac iLon chan che cxpErli mental rrsultc
is thL dth.,rig r nltiion ot singl.. b.jn.1e to ifurrn Joubli bonds.
Furthi rni.ru Lil n ~1i ni tdr1n i oi in -L.' ,-.p ritm n L.,1 reBs l Ls, approximately
-05 kiloLuritLS pl r gram mole of oxygenen r'-dcted, is somnehat closer
tJ the .'alue of the dchyd.lroenition reaction, -55 ki calories pir
gram mol. than it is to the .'alue of the pol..m. ri:ation reactions
or functional grup formations, about -90 ki:localories per gram
mol. This imiplLCe that dehlidrigt- nation of the anphalt is the princi-
pal reaction caking place duringg air blowing, accounting for ovir
half ol the oxygen rract4.d.
According to Gun (37) the aliphatic chains are not dehydro-
gc.nated during air blowing, although naphthenic rings may be dchydrogen-
aled. If this is th.L case, then it appears that only one or possibly
two of the bonds in the naphthenic ring are dehydrogenated. Dchydrogen-
ation of a naphthenic structure to produce an aromatic structure
has a more negative heat of rcacticon than dchydrogenatior, of or.l
one or tro bonds b-cause of resonance in the aromatic rines. SLnCL
this value, 83.0 kiloLalorIts p.r graia molt, is higher than any'
of rh. e:xptrimentdl v\aluLs, th,- ie'sults LItLa:L.e haL aT rum tii..uattor
dois not occur to a v..ry grct t -..tent in dir blu-in,,.
Because th,- fijrmattoi u f an ox.',~ln trojp (e.-.. ptL tLh r ano
anhydride group.) produces a leat of reaction more ineIjti'.e tl.ir,
-90ckiLoc.lori,-os ptr' ram &iol, and bi.cause ithi S typLE of tLi.rLns
must occur because of thi incrcasu in ite oxygen cc'nteii: ot the
asplialt, thLn it must follow that rLe actions 'iIllt .1 riaLiti lIrgL
negative hear oz reaction lLat do not contribute u..-:,g co Lhe
asphalt st ucLurvt take place to a iath r litcdi J L-. .t. tI
The fact that Lther formation Lsults in a less nei.gatiL
heat of reaction than other oxygen group toL-mnactni, unich in Morc in
agreement with the experimental results, lenis suppoir to h the cor.'
that ether bridges are formed in the -jir blo.inrg rL-iction, aIlhough
the experimental results c,.itainly do not justify a definitt stace-
ment on this matter.
In addition it is noted that the heat of reaction values of
asphalt S-62-3, the Gulf Coast naphthenic. are lis.e negative than
those for the other two, indicating that dehydrogenation may bt even
more important in this asphalt than the others. This is not
surprising because the Schweyer-Chipley (9.) analysis of S-b2-3
shows that it has the highest naphthcnic + paraffinic fraction and
the Trailer-Schweyer (103) analysis shoLu, that it has the highest
saturate fraction. hence S-b2-3 has more saturated bonds available
for dthydrogenation. It was also noted qualitatlvtly that this
asphalt consumes more oxygen u.tth a smaller change in Elov proprrcies
than the other tvo, whichh might bu because. dehydrogenation would
haveI less cifect on the \tscosity of the asphalt than polymerization.
B. TicmpcratnurL Effect
The 'lldi.in that th. ILat of r-eatci-r. '.,alues are slightly
less neg.ait in the high t~cmc rature range of 535 to 50) OF. than
liLL lIo' range oi 4oO to -;-5 '. can be attributed to the dccaibo;x''-
lation thit h.s b-i propose by Coppl and Knlotnerue (34) to occur
at highel r teniperatur-s. 0-.t.,dn Irn ls s3 of air blown sa;rples of
S-b2-3 obLtined from runs at difftrt-n te mperaturLS shows that at a
temnpcraturL of 535 to 530 OF. u.hen U.060 pound molos of oxygen
react with asphalt, O.01b pound moles are retained by the asphalt,
while. at a tEnperaturt ot 400 to 473 OF., 0.084 pound moles of oxygen
must react before the same quantity is retained by the asphalt, or
about L0 per cent more, indicating that a greater percentage of the
reacting Lo:ygenr becomes chemtcally bound in the asphalt at lower
temp. raturIs. This contirmns the statement of Goppel and Knontnrus (34)
proposing decarboxylation and the preferential formation of carbon-
carbon bonds over ester bonds in high temperature polymerization.
The decarbox5-laton to form carburn dLoxi.le wouli rEsult in a less
negative heatL of reiaiL'un of about .6 kilocalorits p.-r bram nole
wrightcd by the fraction of oxygen actually IEa'.'n ch sth .stem a'
The significant asphalt-t mnpec3racu interactiun due to cht
behavior of asphalt i-6b2-2, the Last Tc:-:aS i phaltic rTL-tduum, is
more difficult to analyze. The uftfLc is obviously nor ..iircly due
to decarboxylation because che effect is larger chjn -4. kilucal:ries
per gram mole and ducarboxylation is a reaction of rtlacia.-ly iriLor
imporLance. This tenmpcrarurc effict of s-bt2-2 uas also Liticr d by
Ariet (6), who reports that this particular asphalt hia th- highest
energy of activation in the air blowing r-~ation of six jsphalts
that he studied, although he giv'L no c:xpla natLon for ics high.-r
value. The Traxlcr-Schweyer (103) analysis of S-62-2 reveals that
it has the highest asphaltic cunc.,nt of the three asphalis stulicd and
the air blowing kinetic rate was shown co be dependent upon the
asphaltic fraction concentration at a temperature of 550 OF. (6i.
It has also been reported by Goppel and Knotnerus (34) that .aromatic
asphalts are more susceptible to functional oxygen group formation
in air blowing than other types. The asphaltic content of S-o2-2
may be related to a high concentration ot aromatic compounds,
based on the djt3 of Rombcrg, NJesr.ich, and Traxlier (So). Thcrc.iore
the jdcarboxyl Jlaon effect on the hCLc of reactLtoi would be e\:p.cted
mnor for 5-o2-2 and in adn..lirlin Art.-- 's () results inicaLe thia
th I. inLetIL cLnstanLS of asph.iitL with a high asph,iltic fraction
conc.Lnt ire more- susceptible to terampratur.. than thi constants fiLr
oLthr asphalt. The combination of thes,. two phenomena could very
ltkel, be rrsponsiblu for the signific-nt asph.lt-temperarur.,
C. Cat3l'tic ,\ir Blo ing
The results obtiincd for the hCea of reaction of catalytic
air blowing asphalt b-b2-3, the Gull Coast naphthenic residuum,
ITr surriarzed in Table 10.
SUI-M-ARY OF CATALYTIC AIR BLOWL.:I;ZG RESULTS
Catalyst Heat of Reaction,
Aluminum Chloride -71.0
Phosphorus Pentoxide -61.8
The observation that the ditterenC in che hear of reaction
of catalytic air blowing with phosphorus pen'o::tidt and air bl;.oing
with no catalyst ,as nar signi icanc yields tlttt1 in tor:I tiion
concerning the ffectL of this catalyst. Since Hoibrr- ('-) r-porcs
that P 0 increase ihe p-nt.-ratltn by 147 p.,r cent and ch blowing
time by 6 p-r Cent, it obviously cannot be concluded chat the
catalyst has nc. effect on che reaction. However, becauSU thE
magnitude of the heat of reaction is not altered sisnificanct:, by
the addition of P05,o it is improbable that a reaction ot thL type
with a low heat of rc.action is preferenttally accelratLd o.'tr
reaction with a higher heat of reaction or vicE versa. An o::ygen
analysis of a sample of P205 blo..n asphJlt sho'-b that only about
7 per cent of the reacting ox.genn is bound in the asphalt as
corTpared with over 20 per cent in air blowing without a catalyst.
This point, in addition to the grrat change in the softenin. point-
penetration curve shown by Hoibcr6 (44) could indicate that che
formation of oxygen containing groups is preferLntially retarded
in favor of a polymerization reaction, especially by direct carbon-
carbon bond linkage. Since the magnitude of tho' hears of reaction
IIIEf these two reactions are nearly identical, such preferential
cIEatalylss would not be revealed by a change in the magnitude of the
IIeaci oII reII action. Once again, however, the possibility of coipen-
tiniiiiiii prduced by a complex of reactions makes it difficult t to
state anythingg about this catalyst conclusively.
The highly significant differcnce in the heat of reaction
obtained which the use of aluminum chloride, which decr, ascs the
iagritude of the hcdt ,f reaction from -t2 to -71 kilocalories p.r
gram mole indicates that the types ut reactions irth inore ncgati.v
hcts of rcacrions are prefcrientiilly accelrat-ed at the expense of
those ith less negative heats of reaction. Holberg (44) r.portE
that hoe softening point-pencetracion cur.v is altered little by the
use of this catalyst, which would indicate th.at the polynerization
reactions are nor aifecred. An oxygen analysis of a sample taken
during air blowing with AICl3 reveals that 31 per cent oE the reacting
uzygen is retained by the asphalt as compared with about 24 per cent
during comparable air blowing with no catalyst. It is postulated,
[hercforL, that a, prfcrenclal ;jccoleration of oxygen group formation
which has a heat of reaction or about -90 to -100 kilocalories per
gram mole and a retardation of the dehydrogenation reaction which
has a heat of reaction of -55 kilocalorie per gram male is the
probable effect of aluminum chloride since this would result in a
more negative over-all heat of reaction.
U. Sulphurization of Asphalt
The highly significant difference in the heat of reaction
of the asphalt-sulphur reaction due to the type of asphalt demon-
strates that certain reactions play a more important role with
different asphalts in this process as well as in the air blowing
process. Th,. results obtalndJ lor hei hea:a of reaction for Lrh
asphalt-sulphur reaction are aunnarized in Tabl,. 11.
SUIM- RY OF SULPHUFIZATIOI RESlILTS
Asphalt l.atL of R,~action,
kcal gir. nmul
East Texas Residuum, S-62-2 +12 2
Gulf Coast liaphthcnic, 5-62-3 +8 2
WesL Texas-New Mexico, 5-63-3 +9.0
The heacs of reactions for th.e ,ari ous types of isphilc-
sulphur reactions that have been proposed to take place arc given
in Table 12. The same asuaLptions used in the calCulations of thi.
air blowing reactions were applied in this cas3 uith thl ec:'.ptiJor
that hydrogen sulphide rather than water ';as produced. Thi. Individual
calculations are presented in Appendix F. The values are report
in kilocalories per gram mole of sulphur reacting.
It is immediately clear that the ex pLrim. ntal rcs ulcs ar
lower than most of the heat of reaction values lisLtd in Table 12
and so those reactions with a relatively high heat of reaction must
take place to a very limited extent. The dehydrogenation reaction,
which was proposed to be an important reaction in the air blowing
CALCULATED HEAT OF REACTION VALUES FOR PROPOSED SULPHIUPJRZAIOI REACTIONS
Proposed P.tacion Heat o[ Reaction,
keal / gm. ,ol e
DLhydrogination of aliphiLic bond to torm olefin +2u.8
Aromati.:atLon of naphthenic ring to aroria-eic ring +10.9
Direct carbon-c rbon bond lrnnat ion +6.4
Thiol formation +10.4
SulphiJl fonnaLion +10.1
Closure of aliphatic chain to rhioph.nr- type ring +25.9
Dehydrogenation followo-d by sulphide linkage alpha
to double bond +17.4
Dimerization reactions of alkyl aromatics
Aliphatic linkage +6.4
OlefLnic linkage +15.6
Thiophene type linkage +19.1
////////////////////////////// ////////////////i/ii ////////i i/////////////
process has a heat of reaction of 2-.9 kLlocalurrLs pr gr.an m'.le
wirh sulphur, whichh is almost pruhibiti.ecl, high. ThLi is not
unexpected considering Brooks' (0u) report chat olcfina are very,
reactive 'ith sulphur. Furrh-rmjore, the Gulf Coast naphtheric
asphalt, S-02-3, which is expccCEd to ha.,e ihe riose aliphatii bonds
available for chy-drogenation, has the lowest heac of rctcrion of
h'e three asphalts studied. If dehydrogLnarion of parJitins did
occur, it would be expected that the high heat of reaction of this
reaction would make the experimental value for S-o2-3 higher than
the other two.
Although the evidencL indLcaLts that dLhydrcgenation of
aliphatics does not occur, it is quite likely: chatc he d,,h;/rogcn.-
tion of naphthenic rings to aromatic structures, as suggested by che
work of Ruzicka and Meyer (89) and Euzicka, Mlelr, and tlintaziinL (91),
does cake place. The heat of reaction for this r.accion, 10.4 kil-
calories per gram mule, is in the' range of th< e:-.primiLntal results.
Pryor (83) reports that this reaction takes place '-irh man,
naphthenic structures since the aromatic structures torned are not
further reactive with sulphur as .are the olEfins. The air blowing
process heat of reaction results indicated that the dchydrogen.tion
of one or two bonds of a naphthenic ring is one or the major
reactions. In the sulphurization'process, it is likely that the
S same type of reaction takes place, except that the high reactivity
IIIIh^of the sulphur results in all three bonds of the naphtheniL ring
Thi hits of reaction in the formation of thiols and sulphides
.rir in the order of mjgnitudoc of t[hu tperimincal results. but
Tuck.cr's (1_4) finline that only -,bout 15 p r c -rt of the sulphur
r' actin. 'With LhL asphJlt rl,.Lns' chemically bounr in chL asphalt
prohibits ch.si reactions fruo pla3inc more thin a minor rolL.
Tuckr h.-s found, hot.'.'er, thuL the sulphur contenL ot the asphal-
rtnes in,:rases during sulphurizat[ion, speciallyy with lo sulphur
asphalts. This contrasts %ith Gun's (36) finding; thac the ox;g,-n
conLent of asph iltiis decreases during *ir bLoing. and indicact-s
that polyi; rizartion by sulphur linkage may be .more important in
the asphalt-sulphur reaction than pol-mrirzation by oxygen linkage
is in thL air blowing rLaction. 7he magnitude of the heat of reac-
tion values about 10 kilocalornL per grain mole indicates that this
polymtrt-.i'atiun may indeed )occur by the formation of sulphidis.
The fact that the closure of an aliphatic chain to form a
thiophr.ne type ring has a heat of rLaction value much higher than
the e-xperimental r.2sults indicate's that this reaction must take
place to a rath-r limited txtunt, if at all. Tucker's (104) results
bear this out; he has found th3L none of the reacting sulphur becomes
bound in the parafflnic + naphthenic fraction obtained by the
Sch:.Lcy'r-Chipley analysis method (94).
Polymerization by duhydrogenation followed by sulphide
cross linking alpha to the double bond has a heat of reaction of
about 17 kilocalories per gram mole of sulphur reacting, which is
somewhat higher chan the e:..primcntal results of .ibout' 12, 9, and
a ;ilocalorics pr gra-. mole for 5-t2-3, 5-02-J, and S-o3J-.,
respCct v, I;,, buc LE is a r'.'cr: 11.kel possibility nLc..rtheltf s Leciaujs
the tm~iperaturL of the sillphurtzation protLcs is fa'vorble fotOL th.-
reaction, according to Ucsclake (107). ihe form.iTrion oif direct
cjrbon-carbon bonds mintt also be considrcdu v.r/ probable bc-_ause
it is the only reaction with a lowcr heac of reactLon than all of
the expertm'.ntal results. Th.-' c L.. r.atciins OccuLrrine ;sirull-
taneousl/ would have an ov, r-All heat of reaction with a mnuanttorid
in the range uo the .expriLntnal rc.iil a.
The reason for the hLg;il, silnifican'. diffi:r-.ne in the
magnitude of the heats of reati-ion for the three asphalts is difft-
cult to detLenine. The most n-iphthLnic of Cth chrle by c mlpoi Lnt
analysis, the Gulf Coast naphrhcnic, --o2-3, haS th, l0-c3t hjat of
reaction, 8.2 kilucalorics per grum moLL, whilee the most asphiltti,
the East Texas, S-62-2, has th,, hi.hcst, 12.2. This implies that
naphthenic asphalts are more susceptible to polali,,rization b.
carbon-carbon bond linkage and that arrntatic asphalts arL more
susceptible to polymerization by sulphide linkage. Another .-*.plana-
tion could be that Horton's (48) dimerization reaccions of alk.l arorma-
tics take place to a certain extent. Olefinic Linkage or chioph. n
type linkage by aromatic structures would be expected to result in
an elevated heat of reaction value for the more aromatic type asphalts
It should bc LmphaaizLd that th. purpose of this research
w'as to miasurL and report heat of reaction data for the processing
of asphalts. ArtcmpEs to explain r.-chanism front the values of the
h,-at of reaction art valid only w.'hrn s.upplemnntcd by the finlings
of Ari tL () Tuckl r (10-). and Pusor (13), as 1ell os other
in -'E cil gators.
The following conclusions concerning the heat of reaction
of processing asphalt can be dra',n.
I. The asphalt air blowing rE3ction vas found to be .xo-
thermic .ith a heat of reaction varying bEtcxen -61 and -72 ktlo-
calories per gram mole of oxygen reacted, depending upon the typ.:
of asphalt and the blowing temperature. The efiect of th. trpe of
asphalt uas signiEicant at the 99 per cent confidence lii1r, and the
effect of temperature was significant at the 95 p.r cent confidJnce
limit. The extent of completion of the air blowing process 'as
found to have no efEect on the magnitude of the heat or reaction.
2. The results indicated that Jch.,drogenation of single
bonds to double bonds, especLally the single bunds in naphthcnic
rings, may account for over half of the oxygen consumed. ThL heat
of reaction of one asphalt, a GuLE Coast naphthcnkc residuum,
S-62-3, indicated that dehydrogenation is more prominent in this
asphalt than the other two studied.
3. The heat of reaction of another asphalt, an East Texas
asphaltic residuum, S-62-2, was more sensitive to temperature than
the other two. This was attributed to its possible higher aromatic
4. Air blo-ing with one catalylt, phosphorus pentoxide,
proved to have no cffcect on the inmnitude of the h'at of reacrLon,
but air blowing with another, aluminum chloride, d.crearc-d the iugni-
tude of the heat of reaction by 9 kilocalores pl.r grai mole, a
difference that was significant at the 99 pet cent confidence limit.
This was attributed to the catalyst'a prcifrential acceleration of
reactions fonring oxygen containing groups at the expense of
5. The asphalt sulphurtdation process was found to be
endothcrrnc with a heat of reaction vJaring between 8 to 12 kilo-
c.lories per gram mole of sulphur reacted depending upon the type of
asphalt, which wis significant at the 99 per cent confidence limit.
o. The results indicartd that polymerization in the asphalt-
sulphur reaction ma! rake place by both sulphide linkage and by
carbon linkage. Sulphur linkage was postulated to be more important
in this reaction than oxgen linkage is in air bloi.ing polymerization.
7. Dehydrogenation of naphthenic rings was postulated to
takc place in the sulphurization process as well as the air blowing
process; however with sulphur all three bonds of the naphthenic
rings may be dehydrogenated resulting in an aromatic structure. The
formation of stable olefinic double bonds was found to be unlikely.
8. It was concluded that oxygen and sulphur both dehydrogerr
ate and polymerize asphalt. The indications for the differences in
these two reactants lies principally in that oxygen dehydrogenates
only one or two naphthEnic bonds .hLle sulphur d'hy.drogtnatts all
threv, and that polymerization is accomplished more by sulphur
Linkage in the suLphurization process chan by ovgen linkjat In th-
air blowing process.
DR';dINGS OF EQLUIPtENT
EEEE:__ .. :............................
# a ,; 7l
L. 1- I,
a i O -
-- ; *" a c .: .
- s - .-7
- : ,.. ; :
.I,' ,',. "
|-- .. |
J^ -^ ^ I ., _-----------
J / I" j -- -
s -*-. --14 -
S le: 1" -- /4"
Figure-5. Reactor Cover
\I I;" Uijrt ter Il.le -
,- 1/4" B:. 1 c s
o- I r'
O-~: I r .
Diamecr Sceel Rod
Vo'Iage Porcelain Bushing
akel ice Top
1/16" Stainless Steel
S 1" Holes,
O c 0
Insu a tor--- r
4 1/4" Steel Rods
Welded to Pipe Cap
4" Schedule 40 Steel Pipe
4" Pipe Cap
Scale: 1" = 1/4"
Figure 6. Electrostatic Precipitator
S___--1/2" Pipe Couplinbs..
" Schedule 40
2" Schedule 40
1-1/2" Schedule 40
'3/4" Schedule 40
Scale: 1" = 1/4"
Figure 7. Double Pipe Hear Exchangers
Scale: 1" I/4"
S2" Pipe Plug
L- - 2" Pipe Tee
Reducer. 2" to 3e,"
"--3,'4" Schedu'l 40 Steel Pipe
--2" Schedule 0 Steel Pipe
-i _-Bell Reducer, 3" to 2"
S --3" Schedule 40 Steel Pipe
Bell Reducer, 3" to 2"
40 Steel Pipe
2" Pipe Tee
'' Reducer, 2" to 3/4"
Reducer, 2" to 1/4"
1/ /1/4" Pipe Plug
Figure 8. Packed Column
- ........... _
8 1/4." Bolts
Scale: 1" = 1/4"
-1/2" Pipe C.upling
/I8" Statnles StIeel
T'hreadLd Into .
1/4" Lucite Plac.-s
S.. y. I' z 1LI/
Figure 9, Drying Column
0- 4O 0~
I,~ -- .
2 ^' P'pe C~ u f.'Ii;.-
I : .. h i C I, i .,,
Fr I -jr. I. ,- i r- t.- r
_____----- ______- --------li-I/B -
All Holes 1/4" Diameter
Scale: I" = 1/2"
Figure 12. Preheater Cover
I ) i )
Figure 11. Preheater Baffles
(i ) (J
/ \ I i '
( ) ( )
',- i r ,
,' " '.\.. ',
( ) 1 ( .
1 i " _' rtee? 7 i.iot
S C. t -. r-.. er
.l e, .,.. r r '