Volatility of amines used for water treatment in steam generating systems

POYILATHVARIETHBALAKRISHNAN
Atotnic Energy of Canrrdrr Limited, CIznlk Rii.er iVrrrlecrr Laboratories, Chalk Ri\,er, Ont., Canudrr KOJ I J O
Received April 18, 1978

P ~ Y I L A TVARIETH
H BALAKRISHSAK. Can. J. Cheni. 56.2620 (1978).
The relative volatilities of cyclohexylamine and morpholine in dilute aqueous solution have
been measured in the temperature range 150 to 300-C at the corresponding equilibrium
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 180.252.252.83 on 12/04/19

vapour pressure of the solution. Cyclohexylamine strongly prefers the steam phase while
morpholine has a relatively volatility close to that of water. In both cases an azeotrope of the
anline and water, with a vapour pressure higher than that of either component, must exist;
however, with morpholine and water azeotrope formation only occurs above 175°C. The
concentration dependence of the distribution between the steam and solution phases of
both amines is explained in terms of their partial ionization in solution.

P~YILAT VARIETH
H BALAKRISHKAN. Can. J. Chem. 56.2620 (1978)
On a mesure les volatilitCs relatives de la cyclohexylamine et de la morpholine dans des
solutions aqueuses diluees, a des temperatures allant de 150 i 300°C et aux tensions de vapeur
de la solution correspondantc. La cyclohexylamine prcfere fortenlent la phase vapeur alors
que la volatilite relative de la rnorpholine ressernble a celle de I'eau. Dans les deux cas, il doit
exister un azeotrope de I'amine et de l'eau alors que la tension de vapeur est plus klevee que
cclle de chacun des composants; toutefois, dans le cas de la morpholine et de l'eau, la formation
de I'azeotrope ne se produit qu'au dessus de 17YC'. On e x p l i q ~ ~leefait que la distribution entre
For personal use only.

les phases vapeur et liquide des cleus amines depend de la concentration en terrnes de leur
ior:isatioll partielle en solution.
[Traduit par le journal]

Water treatment in nuclear steam generating pholine solutions on the basis of observations in
systems is no\v based predominantly on alkaline boilers. Albrecht and Schuck (3) and Cuisia (4) have
volatile reagents, which neutral~ze acidity in the determined the volatility of cyclohexylamine and
water thereby protecting surfaces from corrosion. morpholine, but in these studies the equilibration
Acidity in the water is usually ca~lsedby carbon was done at 1 atrn pressure. The present work ex-
dioxide from in-leakage of air o r from the decom- tends the studies on cyclohexylarnine and morpho-
position of carbonates in the steam generator. Since line to pressures up to the saturation vapour pressure
carbon dioxide is carried with the steain the carbonic of water corresponding to a temperature of 300°C
acid in the condensate is best neutralized by a reagent (i.e. 8.59 MPa).
which could also be carried into the steam phase,
i.e., a volatile reagent. Thus, common reagents used Experimental
are amines such as ammonia, morpholine, and The distribution of morpholine and cyclohexylamine in
cyclohexylamine. the temperature range of 150 to 300°C (pressure range 476 kPa
to 8.59 MPa) was studied in the apparatus shown in Fig. 1.
A knowledge of the volatility of the reagent is The system consists of a large ( 2 8 L) autoclave with an ar-
important for determining the ability of the reagent rangement to draw samples from the liquid and vapour phases.
t o protect areas exposed to \vet stearn as well as for The arrangement also allows the condensed vapour to be
determining the loss of the reagent during the de- circulated back by natural circulation to the liquid phase in
the autoclave, thereby allowing more mixing between the two
aeration of the feedwater or by blowdown of the phases. However, during the preliminary experiments, the
steam generator water. I n addition, such knowledge same results were obtained with the system operating in the
would also be useful in evaluating the chemistry of dynamic mode with recirculatio~ias when it was operating in
the stearn generator water under 'volatile' treatment; the static mode without the recirculation and consequently
the chemistry will be influenced by the residual amine the latter mode of equilibration was used in the later experi-
ments. When operating in the static mode, the rising portion of
concentratioil in the water. the vapour line in the apparatus was made as short as possible
There have been some studies on the volatility of to rnininlize any error due to refluxing.
amines. Jones (1) has determined the relative vola- For each experiment 4 L of demineralized water was put
tility of ammoilia in water in the concentration in the autoclave. Nitrogen was bubbled through the water for
about 1 h to remove dissolved air. Sufficient amine to give
range 2 to 200 pg NH,,'kg water, in the temperature approsilllately the desired concentration (mg amine/kg water)
range 150 to 325°C. Berk and Nigon (2) have dis- was added to the water and the autoclave closed. The gas
cussed the volatility of cyclohexylamine and mor- pressure over the solution in the autoclave was reduced hy
 BALAKRISHNAN 262 1

perature for 4 h and then 20 to 100 g sample of the solution

phase and 10 to 50 g sample of steam phase were taken via
valves 1 and 2, respectively, for determination of the amine
molality in each phase. The size of the sample depended on
the amine content in the sample. The sample size was kept to
the minimum and the sample was drawn slowly (over a period
of 10 n ~ i n )to minimize the perturbations in the equilibrium.
In the preliminary experiments, san~pleswere taken from the
autoclave at 2-h intervals over a period of 24 h. Analyses of
these samples for amine showed that a steady state with
respect to the amine distribution was attained within 4 h. In
view of the precautions taken to nlinimize perturbations in the
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 180.252.252.83 on 12/04/19

equilibrium during sampling, this steady state should be close

to the equilibrium.
The amine molality in a sample was determined by poten-
tiometric titration of a suitable aliquot against standard
hydrochloric acid, with care being taken to exclude air during
the analysis by bubbling nitrogen through the solution.
Values for the relative volatility, defined here as the
(dimensionless) ratio of the amine molality in the steam
sample to that in the water sample, were calculated.

Results and Discussion

?&
VALVE 1

SOLUTION PHPSE
T.1
VALVE 2

VAPOUR P H A S E
The relative volatility values at different allline
concentrations (mg amine/kg water) and a t difl'erent
temperat~lres are given in Table 1 for cyclohex-
ylamine and in Table 2 for morpholine.
For personal use only.

SAMPLE SBMPLE

FIG. 1. Apparatus for volatility studies at high tempera- The relative volatility of cyclohexylarnine was
tures. always considerably greater than unity, i.e. the
pumping off the gas with a vacuum pump for about 15 min. amine molality in the vapour phase was much
The autoclave was then heated to the desired temperature. higher than that in the liquid phase in the tempera-
Twice during the heating, i.e. when the temperature just ex- ture and concentration ranges studied. The relative
ceeded 100'C and when the experimental temperature was
reached, some steam (equivalent to e 100 g water) was blown
volatility decreased as the temperature increased in
off via valve 2 to assure complete removal of gases from the the range 175 to 300°C, but increased as the concen-
system. The system was maintained at the experimental tem- tration increased.

TABLE1. Cyclohexylamine: relative volatility and distribution coefficient

Tempera- Water phase Relative Neutral Distribution

ture amine conc. volatility K, K,,. fraction coefficient
("C) (mg amine/kg water) (RV) (rnolikg) (molZ/kgZ) (a) (Rv/a)
 CAN. J . CHEM. VOL. 56, 1978

TABLE
2. Morpholine: relative volatility and distribution coefficient

Tempera- Water phase Relative Neutral Distribution

ture amine conc. volatility K, Kw fraction coefficient
('C) (mg aminelkg water) (RV) (mollkg) (mo12/kgz) (a) (RV/a)
158 186 0.86 5.496~ 2.685 x l o - ' 2 0.953 0.90
158 21 0.72 5 . 4 9 6 l~ o 7 2 . 6 8 5 10-l2
~ 0.868 0.83
160 448 0.88 5.833 x 2.793 x lo-'' 0.970 0.90
180 436 0.98 1.034~ 3.908~ lo-'' 0.973 1 .01
180 179 1 .00 1.034~ 3.908~ 10-l2 0.958 1.04
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 180.252.252.83 on 12/04/19

182 25 0.98 1.091 x 4.018 x 10-l2 0.893 1.10

200 436 1.01 1.767~10-~ 4.989~10-'~ 0.977 1.03
200 199 1.12 1 . 7 6 7 ~ 1 0 - ~ 4.989~10-" 0.965 1.16
203 21 1.11 1.911~10-~ 5.140~10-'~ 0.900 1.23
230 414 1.31 3.717~ 6 . 1 3 8 ~10-'' 0.982 1.33
237 20 1.30 4 . 3 7 2 ~ 1 0 - ~ 6 . 2 6 6 ~10-l2 0.925 1.41
25 1 20 1.22 5.991 x 6.353 x lo-'' 0.935 1.30
255 185 1.33 6.542~ 6 . 3 3 9 ~lo-'' 0.979 1.36
273 331 1.30 9.563~10-~ 6.026~10-'~ 0.987 1.32
28 5 22 1.32 1.217~10-~ 5.636~10-'~ 0.959 1.38
300 409 1.33 1.624~10-~ 5.000~10-'~ 0.992 1.34

The relative volatility of morpholine, at tempera- F o r electroneutrality,

tures lower than about 175'C, was slightly less than
For personal use only.

unity. With increasing temperature, relative vola- [6] [BH"] + [H'] = [OH-] = K,/[H']

tility increased and tended t o approach a limiting

value of about 1.35.
The effect of amine molality on the relative [71 (1 - a ) B , = K,/[H'] - [H']
volatility may be explained on the basis of the where K , is the ionic product of hvater. Froin eqs.
limited ionization of the amine in the solution. The [4] and [7J
equilibriun~involved is
[11 BH' B f H+
where B is the base, cyclohexylamine o r morpholine, from which cr may be evaluated. The distribution
and BH' the conjugate acid, cyclohexylammonium coefficient K, for the amine may then be obtained
o r morpholinium ion. Only the neutral (uncharged) from the relative volatility values using the relation
amine is expected to go into the vapour phase. Thus,
[lo] K, = relative volatility/x
the equilibrium for the distribution of the amine in
the two phases would be F o r evaluating a the ionization constants, K,,
determined by Mesmer and Hitch ( 5 ) were used. The
values for ionic product of water, K , , were chosen
The equilibrium constant for reaction [2], i.e. the from the 'best' set of values listed in the review
distribution coefficient, is article by Olofsson and Hepler (6) on the thermo-
dynamics of ionization of water. The calculated
values of a and values for K, and K, are also listed
[B],,,,, may be assumed t o be the total amine mola- in Tables 1 and 2. Figures 2 and 3 show plots of the
lity in steam. The molality term [B],,,,, may be distribution coefficient vs. temperature for cyclo-
calculated from the total amine molality, B,, in the hexylamine and morpholine respectively at pressures
water phase using the equilibrium constant K , for correspoiiding t o the vapour pressure of the solution
- -1
reaction 11 a t the temperature. The figures show that the de-
[BILH'I - [H+, scription of the system in terms of a distribution
C4I K, = [BH'] 1-a coefficient, K,, for the equilibrium between the
where a is the fraction of the total ainine that is neutral amine in the water phase and the amine in
neutral, i.e., the steam phase is satisfactory. The deviation to-
wards lower values for the distribution coefficient in
the case of cyclohexylamine a t the concentration of
 BALAKRISHNAN 2623

x
AIAISE C O N C E N T R A T I O N P R I O R TO
EOUI L I B R A T l O k
observed volatility behaviour then indicates the
o z 6 0 mg a m i n e l k g WATER existence of a n azeotrope of higher vapour pressure
-
5
LL
Iro :2 5 rng arnine/kg H A T E R than that of either pure component, Lvater or the
H

amine. The azeotrope formation lvould be at some

concentratioll of amine higher than those studied.
The azeotrope must exist throughout the range of
temperature and pressure studied (temperature :
150 to 30OCC, pressure : the saturated vapour pres-
sure of the solution corresponding to the tempera-
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 180.252.252.83 on 12/04/19

ture). Since the vapour pressure of water increases

rapidly with temperature, the decreasing trend in
TEilPERATURE "C
the distribution coefficient observed at temperatures
FIG.2. Distributioll of cyclohexylainine between steam and higher than 175°C might be reflecting the nature of
water. the cyclohexylamine-water phase diagram, such as
a shift in the azeotrope cornposition towards higher
mole fraction of water.
Morpholine is not known t o form a n azeotrope
with water at the low pressures at I\ hich it has been
studied. The observation of a distribution coefficient
less than 1 (Table 2 and Fig. 3) at temperatures
lower than 175°C (and correspondingly lower pres-
sures) is thus consistent with the available informa-
-. 4 2 0 ~g
For personal use only.

0 allne/*g ilATER
tion. However, with the temperature increasing
0 z 2 0 0 ng snipe k g WATER
above 175"C, along with a corresponding increase in
A 3 2 5 mg a m ! n e / k g H A T E R
a the pressure, the distribution coefficient beconies
I I I I I I
0
150 175 200 225 253 275 300
greater than unity and eventually levels off at a
TEMPERATURE " C
value of 1.35. This behaviour may be explained by
FIG. 3. Distribution of morpholine between steam and the existence of a n azeotrope of higher vapour
water. pressure than that of either morpholine or ~vater,at
temperatures higher than 175'C.
e l 0 mg aminelkg water before equilibration is
probably due to the effect of residual carbon dioxide
in the autoclave during the experiment. The residual B. M. Kent and D. P. Martin. both of Chalk
carbon dioxide nlay effcctivcly bc lowcring thc vola- River Nuclear Laboratories, rendered valuable as-
tility of the amine, the effect becoming more signifi- sistance in carrying out the experiments. I thank
cant a t lower concentrations of the amine. Also, at them.
lower concentrations of aminc, the detcrminatio~lof
1. M. E. JONES. J . Phys. Chem. 67, 1113 (1963).
the amine molality by the titration method becomes 2. A . A . BERKand J . NIGON.U.S. Bureau ofMines. Technical
more difficult and less accurate. Paper 714 (1948).
The distribution coefficient of cyclohexylamine 3. S . H . ALBRECHT and J . J . SCHUCK.Mater. Perform. 15, 23
was observed t o be greater than unity at all tempera- (1976).
tures (Table 1 and Fig. 2). Based on data available 4. G . C . CUISIA.Mater. Perform. 16, 21 (1977).
5 . R. E. MESMER and B . F. HITCH.J . Solution Chem. 6,251
for temperatures up to 134'C, the vapour pressure (1977).
of pure cyclohexylamine may be expected to be lower 6. G. O L O F S S O
and
~ L. G . HEPLER. J . Solution Chern. 4,127
than that of pure water at all temperatures. The
 This article has been cited by:

1. S. Yu. Suslov, A. V. Kirilina, I. A. Sergeev, T. V. Zezyulya, E. A. Sokolova, E. V. Eremina, N. V. Timofeev. 2017. Complex
amine-based reagents. Thermal Engineering 64:3, 237-241. [Crossref]
2. Malgorzata M. Lencka, Jerzy J. Kosinski, Peiming Wang, Andrzej Anderko. 2016. Thermodynamic modeling of aqueous systems
containing amines and amine hydrochlorides: Application to methylamine, morpholine, and morpholine derivatives. Fluid Phase
Equilibria 418, 160-174. [Crossref]
3. Andrey V. Plyasunov. 2012. Values of the Krichevskii Parameter, A Kr , of Aqueous Nonelectrolytes Evaluated from Relevant
Experimental Data. Journal of Physical and Chemical Reference Data 41:3, 033104. [Crossref]
4. Andrey V. Plyasunov, Everett L. Shock. 2003. Prediction of the vapor–liquid distribution constants for volatile nonelectrolytes in
water up to its critical temperature. Geochimica et Cosmochimica Acta 67:24, 4981-5009. [Crossref]
5. P. V. Balakrishnan. 1988. Liquid-vapor distribution of amines and acid ionization constants of their ammonium salts in aqueous
Can. J. Chem. Downloaded from www.nrcresearchpress.com by 180.252.252.83 on 12/04/19

systems at high temperature. Journal of Solution Chemistry 17:9, 825-840. [Crossref]

6. C. L. Sharma, M. S. Islam. 1986. Characterisation of Mixed Ligand Complexes of Cobalt(II) and Nickel(II) with Diphenic Acid
As Primary and Heterocyclic Bases as Secondary Ligands. Synthesis and Reactivity in Inorganic and Metal-Organic Chemistry 16:4,
553-563. [Crossref]
7. E. A. M. Wetton, G. G. Lewis. Paper 99. High temperature pH calculations for solutions of volatile amines 355-361. [Crossref]
8. Norma H. Piacquadio, Miguel A. Blesa. 1982. The interaction of morpholine with metal ions—I. Polyhedron 1:5, 437-440.
[Crossref]
For personal use only.