ASALYTICA CHIMICA AC-l-A 405

AN Al’l’Al;tA’I‘US FOR SIMUL’I’ANEOUS GAS EVOT,UTION ANALYSIS

AND MASS SPEC’I’ROMETRIC ANAT,YSIS

f3cczrusc of the wick variation in the \vuig:llt-lass curves tddainctl by tlicrmo-

gravimutry, Roc;r:,Its, Y;\st:tk\ A?;t.) %kssL introducccl the t~~~llt~i~~~~~ r)f gas cvidutiott
analysis (lincnr pyrolysis). !Vith this tccfinicfuc, it srnnll amount of sample is pyrofyzecf
in a dynamic Idrim or air atmc)sphorc, ztt it litwnr heating rntc, ancl the c~~lvccl
pyrolysis g:EIza:s arc cletectcd, ils a function of time or tcmpcraturc, by ;L tllcrnial
conductivity clctcctor. A similar apparatus h;ti IIWII dcscrild by \fr\ssj\l.l.c~~ for
studying tlrc tllcrm;ll clcgraclation of pc~lytncric materials.
‘I’hc conlbination of gas evolution amtlysis (GISR) with other tiicrrnal tcclini-
clue?; SIX-~ :W cliffcrcntitd tficrmat analysis (u’t’Al), tilcrr-Ilngr~rvin~ctric: i~n;d~~is (IX.%)
and so on, \vf)ulcl bc cxpectud to yiolcl cspc&tlly uscfut results in that the zilrlvnntagcs
of e;rclk tcchniquc could bc utitizccl. As 8 result of this, SC’VctYLl D’i’A-<;EA apparatus
ticsigns llil,VC been clcscribctl~~~, as well as tlie nli~nUf~KAure of cotnmcrciul instruments.
‘I“lle coml~inccl tccllniqucs allow iL more complete intcrprctatic0l of tllc DTA curve,
such as distinguishing bctwccn ;L phase transition ant1 IL dccom~~osition reaction
involving 8 g;wwus product, and so on . An cxccf1e11t cxamplc of tl1c US<!of the u-I-n.-
(;EA tctrhniquc has bcun dcscribcd by WI:,XI~LAS~~~~ASI> STIJR~~.
One itn]~r~rt~Lnt clis~~~v~Lrlt:L~~still pcrsid in the sinltllt~~~l~~~~lsDTA -GEA
tcclmicltic, howcvcr. Tllc C;EA curve intlicittcs that a g;kscr>tls prctduct is cvolvctl but
cannot clctermine its cotnpcdtion. A solution to this prohlcm has been to puss the
evdvcct gases into a gas chromatogr~tpli equi~)ped tvitlt ;L 5uitnblc scparatiort cdumn.
Such a method is cumhcrsotne in that sevcrnl cliffcrent columns arc somctitncs
rccluired for one iU~;llySiS and il conzplctc CktCrIIIiniLti011 cannot bc m:Klc 011 one
pyrolysis cxpcriment. A more rcasonnble appruacl~ to the nndysis of the GISA
products is to lcad the cvolvcci product grtscs directly into a mass spectrometer.
Since most crf ttic evttlvecl gases are tndcculcs having R molcculnr weight under x00,
~7 rather simple and inespenkvc mass s~>ectronretcr may IX empfctyccl. Sucli an
a~q~arntus is clcscribcct here and tvill bc clcsignated ils ;t gas evolution analysis (GEA) -
mass spectronietric wwlysis (MSA) apparatus. Not only dots tllc sitnultaticous
GEA-MSA technique have definite advantages over analysis by gas chromate-
graphic methods, but it cnnblcs the sample to be pymlyzccl at atmospheric pressure
in an inert or other atmosphere instead of the ift uactm conditions of tlic usual maSs
spectrometric tccfinirjues~~. “I’hc i7~ ~tactt~ conditions of pyrolysis arc difficult to
correlate, at times, with thermal propertics obtdnecl by other metliocls.

A nnf. Cliirr8. A&z, 32 (x@s) 405-4x0

 W. W. WJ3SDJaANDT, T. &I. SOUTHERS

A scllem;rtic cliap-am of the apparatus is given in Fig. I, while the pyrolysis

chamhcr is illustrated in Fig. 2.
l’hc (;I% part of the apparatus crmsistcd of ;I pyrolysis &amber, S, furnace,
FU, furnncc tompraturc controller, Cl, snmplc temperature recorder, X32, thermistor

MS-IO

H
-
l!El 0
0
0
RI

rc 0.

t He out

A rral. Cirirh. Ada, 32 (1965) 405-410

APPARATUS FOR SIMULTANEOUS CEA ASD MSA 407

thermat conductivity cell, *IT (GoXv-Mae Instrument Company, Madison, NJ.,

Model gG80), bridge circuit, C2 (Gow-Mac design), and a. S-Y recorder, R3. Helium,
from a cylinder and pressure regulator, is passed through the reference siclc of the
thermal conductivity cell at a flow-rate controlled and measured by a flowmeter, FL,
into the pyrolysis chamber, and then into the sample side of the cell. The gas emerging
from the cell passes throu@x a “T” connection which contains the end of a 4-ft. length
of 0.013-ID stainless steel capillary tubing, CA. The other end of the tubing- is
connected to the mass spectrometer inlet system \vhieh also contains a nxolecular
leak, L. A portion of the gas stream is thus introduced into tile spcctromctcr while
the remainclcr is pumped out to the atmosphcrc by a nxcchanical pump at P. The
gases are then analyzed by the Associated Electrical Industries Ltd. Model MS-IO
mass spcctroxnetcr (distributed in the U.S.A. by Picker S-ray Corp., White Plains,
rU,Y.), the output of which is clisplayud on rccorclcr ICI.
Tl~2 pyrolysis chamber consisted of a Pyrex glass tube, 17 cm long by 2.5 cm
in diameter, which was ttlrminntecl on both ends by a 2.5-cm inner diameter “0’‘-ring
joint. ‘T\vo “0”-ring joints, identical to the glass joint, were machined from nluxninum
and attached to the &ss joint by xxx&al clamps, All metal tubing connections and
internal thcrxnocoul~lc connections were made by Conas connectors fastened ixlto the
aluxninum joints. ‘Tllc glass tube was wrapped tvitlx asbestos p;ll~~r, wound with
enou& Niclrromc hoatcr wire to give a total resistance of about 25 ohms, and then
insulated with additional asbestos IX\pC!r to give a layer about 0.8 cm thick, The
sample block consisted of ~1 aluxninurn cylinder, r ,7 cm in diameter and 2.5 cm in
lcxtgth, containing a glass cup srrmplc hulder, o.8 cm in diaxnetcr and 1.5 cm in Icngtlt,
‘The entire sample holder assembly rested on a two-holed ceramic insulator tube (3
mm diameter) which held a Chromcl-Alumcl tltcrmocouplc. Thus, sample tenlpcra-
turc. not furnace temperature, was nlcnsurcd. ‘i‘lic pyrolysis furnace tcmpcraturc rise
was controlled by a motor-driven vuriable transformer, ~vhicli has previously hccn
clescribcd7.
The metal tubing ft oxn tlte pyrolysis chamber to the thermat conductivity cell,
the thermal coxrductivity cell, and the capillary tubing, were all maixltaincd at about
80” by means of external heating jackets.

PYocsdtrYe for fiyrolysis

From 8~x20 mg of sample was plwxi into the sanq>tc ectxxtnincr ax& the
pyrolysis chamber nsscmblcd and xnutlc gas-tight, Helium was flushecl tlxrou@ the
system at a flow-rate of 200 ml/mixi for about 5 min. During this tixne, the thermal
concluctivity cell bridge circuit was allowed to stabilize at a bridge current of 8 mA
and a residual gas mass spectrum wks clctcrminecl over the mass range to bc studied.
The mass slxctromctcr cnuld be s~antwd ~~t~tinu~usly from w/o x2 to 45 or r8 to IOO
at a rate of one complete scan in 3.5 xnin when a ro’o- ohm clectromctcr tube and
rapid sc;lf\ motor were employed.
‘I’lic flow-rate was then lowcrcd to 50 ml/miri and the heating cycle of the
furnace bc~un. Usually, a furnace heating-rate of 4-G”/min wns employed, but this
coulcl be varied at will. The licetinfi-rate curve was recorded on the strip-chart
recordex, Rz, while the gas evolution curve was recorded on S-Y recorder R3,
Several modes of operation of the output of the matis spectroxnetcr arc possible. The
various ?tt/c peaks can be recorded ~(~ntinLxously 011 recorder Iiz at a rate of one scan

And. Chim. Am, 32 (.rg@j) 405-410

408 W. W. \VEh’I~I.AKI~T, T. M. SOUTHEIth’

every 3.5 min. Or, the sl~cctromctcr can be set on one ?tz/e value and the output
recorclctl as a function of time on recorder 1x1 or of temperature on rccordcr 113.
Usually, the latter method WIS cmployccl after an initial run had been performed to
cletermine the comllosition of the cvolvecl gases in the GEA curve peaks. If several
different products were evolved during the pyrolysis, multiple runs were made rach
at a cliffcrent ns/e vnluc.

The utility ot tllc apparatus is illustrated by studying the thcrmd dissociation

of [Cu (N &)r;lSOz~ - f-Ml. A~a_W has previously shown that igt VBCM, [Cu(NH&]S04
hScS (JlIc II-de Of zumrmni% per mole of comldcx heginning at 30” and a sccc)ncl mole
at 55”. At still higllcr tcmpcratures, grcntcr than go”, decomposition of Cu(N H&SOs
OCCU~Rwit11 tllc gradual formation 01 ;L dark-grcy to brown product, accompnnicd by
the sublimation of ammonium sulfate In ;I more recent investigation, l’I:ExI%l~~
postulatetl the following dissociation sccluencc:

,210°
[Cu(NH:,)~:JSO~ . I-Is0 = CuS0.r . zNF13 . IInO -+

CuSOlr . zNH:, =% CuSO~ . NHn -=% CUSO,I

From the crystal structure of [Cu(N Ii:~)+O,~ Hd..), as rcportctl by M~zr 10, it is highly
l

unlikely that the ttlcrmal dissociation proceeds as given by Pif:r;xi:I’1<~‘.l’lic complex

consists of layers of Cu(NH&- “+ ior>s l~onclcd to layers of sulfate ion tetrahcdra l>y wa-
ter molacules. Thus, it is ulllikcly that the ammonia shoulcl IX evolved before the water.
The gas evolution curve ant1 the mass spcctromctcr curves :ire given in Fig. 3.
The GBA curve consists of 6 maxima with peak tempcrnturcs of r45”, x75”, zS5”,

.E _ -mm..
MSA

200 300 400

temp. *C

Fig. 3. GEA and hlSA curves of [Ch(NI-I~).~~SO.~ + I+90; ilciltillg rntc of G.oO/nlin; snmplc size of
80 mg ; helium flow-rcltr?of 50 ml/min.
APPARATUS 1’01~ SIMULTASEOUS GEA AXD MSA 409

355”. 380”, and &o”, respectively. B-v itself, the GEA curve is impossible to interpret
nltl~ougll “intelligent gue~scs” may bc made. All that can be definitely concluded is
that each peak reprcscnts the evolution of a gaseous product which may be water,
ammonia, or other substances.
The MS.4 curves, lmwcvcr, enahlc a complete interpretation to be made of the
thermal dissociation process. The first peak, at x45”, is clue to the evolution of water.
This peak was recorded by a pyrolysis scan at a W/Cof 18. The next 3 peaks, at 175”.
29o”, and 360”, are clue to the evolution of ammonia, as revcalcd by a pyrolysis scan
at ‘a W/E of 17. No correction was made for the contribution of water to tl1e atnmonia
peaks. The 3S2” and 460” peaks were due to the evolution of .;ulfur dioxide; these
were causccl by the ciecompot;ition of anhydrous copper sulfate to CuSO4 * CuO11.
Thus, it is seen that tl1e interpretation by PPEIF~SR~ is entirely erroneous. The
tl1crmal clissaciation of [Cu(NH3)4]SO4 0 H& takes place according to the scclucncc:
[Cu (N H:,)4]S04 - HzO(s) + I:Cu(NH:+,]SO4(s) + HzO(v)
(Cu(NI-r3)4]SO 4 (.)
s --j. Cu(NHn)n_SO.r(s) -t_ zNHn(g)
Cu(NHn)zSO 4 (.)s --t Cu(NH3)SO&) + NI-I:%(g)
Cu(NHa)SO.r (s) --f CuS0.a (s) + NH:t (g)
CllsO.+) + cue * CUSOJ (s) $- so?(g) + $0,(g)
Not a &Teat amount of significance can bc placecl upon tl1c peak maxima temperatures
since they are dq~cnclent on the pyrolysis apparatus cxperimcntal parameters such
its sample size, carrier gas flow-rate, furnace heating-rate, and so on1.
Thus, the use of the GEA-MSA technique enables the exact intcrprctation
of thermal decomposition reaction involving low molecular weight products. It is
unlikely tl1at quantitative clata can bc obtainccl because of the difficulty in regulating
the pressure in the mass spectrometer. However, quantitative data are easily obtnincd
by other chemical methods sucl1 as sample collection on various absorbents and
others. The qualitative identification of gas mixtures is, however, easily carried out
with this apparatus. ‘I’here appears to bc no reasc>n why the MSA technique could
not bc couplccl witl1 other thermal mcthocls such as ‘WA, lXA, high temperature
X-ray, pyrolytic mctl1ods, and others. Such an attempt has been mntle wit11 DTA*“,
but tl1c sample was pyrolyzccl in ZJNCILO causing a reduction in the resolution of tl1c
DTA peaks.

The support of this work by the U.S. Atomic Energy Commission and tl1c
Robert A. Welch Foundation is gratefully acknowlcdgccl.

An apparatus is describecl whicl1 enables the g;w evolution analysis (GEA) and
mass spcctromctric analysis (MSA) curves of a sample to be recorded simultaneously.
The sample is pyrolyzed in a chamber in a dynamic helium atmosphere. The cvolvccl
products arc detected in the helium gas by a thermal conductivity cell which results
in the GEA curve. An inexpensive mass spectrometer is used to monitor the helium
gas stream whicl1 gives the MSA curve. The advantages of the apparatus over other
recent techniques are given.
4x0 W. W. WENDLAXDT, ‘r. h1. SOUTHERN

1~s xutcurs chhivcnt un aplxu-cil permettant l’cnrcgistrement simultand des

courhcs oI,tenucs par I’analysct clc gaz pyrolysh, ainsi clue celles dc l’analyse spectro-
m&rique dc masc. Lcs procluits form& sont cl&e&% h l’aidc d’une ccllule de con-
ductivitd thcrmiquc. I,‘&hzlntillon cst pyrolysd dans unc atmosphbe d’l~Clium. On
clchrit Its :tv:mtitges de cet appareil, par rapport h d’autres techniques r&xntes.