MARCH, 1936 INDUSTRIAL AND ENGINEERING CHEMISTRY 375

From the trend of the data on salt effect by Myers and Baker
( 5 ) , it may be assumed that a t p H values below 3.35 the
addition of a slight amount of salt will bring about a decrease
in strength of jellies. The lower the p H the more rapid will
be this decrease for a like amount of salt. I n the comparison
of the strength of jellies made u p a t the same p H from the
first and second extractions from the cranberries, it may be
said that the total cation Concentration is higher in the first
extraction, and that coagulation and precipitation or pre-
gelation tendencies are greater. Therefore, the strength of the
jelly would be lower. According to Spencer (11) a sodium
salt decreases the jelly field on the acid side. It may be as-
sumed that the jelly strength would be similarly affected by
the potassium salt. It is well known that potassium acid
tartrate has this effect on the formation of jelly from juice
extracted from overripe grapes.
The differences in jelly strength between the jellies from
the lemon and the cranberry pectins may be partly due to
the differences in actual pectin present. A determination
showed 75.76 and 69.28 per cent pectic acid present in the
0.5 per cent solutions of the lemon and cranberry pectins,
respectively. This difference in pectic acid is not considered
as important in jelly formation as the degree of polymeriza-
tion of the galacturonic acid ( 7 ) .
The relative concentrations of basic ions (calcium, mag-
nesium, potassium, sodium) in the pectin from the lemon and
the cranberry may account for part of the variations in jelly
strength of jellies from the two sources. This phase was not
within the scope of the investigation. Consequently, analysis
of the pectin ash which would bring out these relationships
was not made.
Acknowledgment
The p H measurements were made by G. M. Gilligan.
Literature Cited
(1) Baker, G . L., Food I n d . , 6 , 305 (1934).
(2) Cox, R.E., Ibid., 5, 348 (1933).
(3) Fellers, C. R.,Am. J. P u b , Health, 23,13 (1933).
(4) Morse, F. W., J . Biol. Chem., 81, 77 (1929).
(5) Myers, P.B., and Baker, G. L., Del. Agr. Expt. Sta., Bull. 144
(1926).
(6) Ibid., 149 (1927).
(7) Ibid., 187 (1934).
( 8 ) Nelson, E. K., J . Am. Chem. SOC.,49, 1300 (1927).
(9) Olsen, A. G.,IND. ENG.CHEM.,Snal. Ed., 6, 143 (1934).
(10) Olsen, A. G., S. Phys. Chem., 38,919 (1934).
(11) Spencer, Gene, Ibid., 33,2012 (1929).
(12) Tarr, L. Wr., Del. Agr. Expt. Sta., Bull. 142 (1926).
R B C ~ ~ I VSeptember
BD 23, 1935. Presented before the Division of Agricul-
tural Chemistry at the 90th Meeting of the American Chemioal Society,
6an Francisco, Calif., August I9 to 23, 1936. Publication approved by the
Director, Delaware Agricultural Experiment Station.

Specific Heats of Sodium

Hydroxide Solutions
JOHN W. BERTETTII AND WARREN L. McCABE

H EAT balance calculations of processes in-

volving the mixing, diluting, or concen-
trating of solutions must be based on the
specific heats and heats of dilution of the solutions. It has
been shown (4) that such thermal data are conveniently
University of Michigan, Ann Arbor, Michigan

cent and a t an average temperature of about 17.5" C.

(63.5' F.);by Tucker (14) over a concentration range of 16.5
assembled in the form of an enthalpy-concentration diagram, to 47.4 per cent and a temperature range of 18.9"to 24.0" C.
or Merkel chart (6),from which items for heat balances are (66.0" to 75.2" I?.); by Pratt (6)over a concentration range
readily taken. of 0.6 to 16 per cent and a temperature range of 0" to 33.3" C.
When, however, the construction of enthalpy-concentration (32" t o 92" F.); by
charts for common commercial chemicals is attempted, it is Richards and associates
soon found that the necessary thermal data are usually lack- (7-11) over a concentra- New d a t a , o b -
ing. As a start towards remedying this lack of data, this t i o n r a n g e of 0.139 to t a i n e d primarily
paper presents new experimental results on the specific heats 27.8 per cent and a tem- for industrial use,
of sodium hydroxide solutions over a concentration range of perature range of 15" to are reported on the
4.08 to 51.15 per cent of sodium hydroxide by weight, and a 20" c). (59" to 68" F.); specific h e a t s of
temperature range of 3.1" to 88.6" C . (37.5O to 191.3' F.). and by Gucker and
When combined with data on heats of dilution, which have Schminke (3) over a con-
s o l u t i o n s of so-
been presented in another paper ( I ) , the new data allow the centration range of 0.158 dium hydroxide in
construction of an enthalpy-concentration chart over the to 8.157 per cent a t a tem- water. The meas-
same ranges of concentration and temperature. peratureof 25°C. (77°F.). urements were ob-
The work was done entirely from a practical point of view.
No attempt is made to develop or test theories of solutions or Experimental Pro - tained by the adia-
batic method and
to present the data on any but a n empirical basis. Extreme cedure
accuracy was not expected, but the results are believed to be The adiabatic method cover a concentra-
of adequate precision for all engineering applications. developed by R i c h a r d s tion range of 4 to
(11) was used. The calo- 51 weight per cent
Previous Work r i m e t e r p r o p e r , sur- sodium hydroxide
rounded b y a ('sub-
Specific heats of sodium hydroxide solutions were reported marine" was totally im- and a temperature
by Thomsen ( I S ) over a concentration range of 1.1 to 23 per m e r s e d in a bath; the r a n g e of 37" t o
1 Present address, Standard Oil Company of Indiana, Whiting, Ind. temperature of the bath 191" F.
 376 INDUSTRIAL AND ENGINEERING CHEMISTRY VOL. 28, NO. 3

V/d/,l 1
I 1I IIIIIIIIIII IIIIII r 11

FIGURE
1. WORKING
CALORIMETER
2. COLDCALORIMETER
FIGURE

was maintained, as nearly as possible, equal to that of the completely immersed in a bath, K, which is contained in the inner
calorimeter. The heat input was determined by measuring can, S. Can S is lagged by 1 inch of felt, T, which, in turn, is in-
side outer can V . Reciprocating stirrers, C and D, were operated
the voltage and current of an electric heater and the time of a t a rate of sixty double strokes per minute. Each stirrer had a
heating. Corrections for heat of stirring were found by short Formica section, g, to reduce heat conduction. Chimneys
measuring the temperature drifts before and after the heating are shown at M . The quantitative heating element, E, was simi-
period. lar in design to that used by Gucker and Schminke ( 3 ) . An auxil-
iary heating element, F , was used to bring the calorimeter ra idly
to temperaturewhen operating at elevated temperatures,and Keat-
Apparatus ing coil 0 could be used to heat the bath rapidly to the working
The calorimeter was so designed that with a slight modifica- temperature. Adiabatic control to within *0.02O F. was made
possible through use of a differential thermoelement of twenty-
tion it was possible to measure either specific heats or heats four pairs of wires, &, directly connected to a sensitive galvanome-
of dilution. All parts which came in contact with the solu- ter, and the use of electrolytic heating (Id,2). The thermo-
tions were of pure nickel. All thermoelements were of No. element junctions were insulated by a mixture of ultramarine and
40 copper and No. 30 collodion, and were im-
mersed in mercury to re-
c o n s t a n t a n wires and r duce lag. In connection
were c o n s t r u c t e d and with electrolytic heating it
used according to the --A 1 1 was found that when the
r e c o m m e n d a t i o n s of submarine itself was used
as an electrode, true adia-
White (16). batic conditions did not
Figure 1 is a diagram prevail when the adiabatic
of the calorimeter: bath thermoelement indi-
cated zero temperature
Vessel A has a capacity difference. The sub-
of about 650 cc. (22 fluid marine was a c t u a l l y
ounces) a n d contains slightly warmer than bath
the solution undergoing K. This difficulty was
investigation. The cover overcome by surrounding
is fastened to the flange, the s v b m a r i n e with a
a,by means of the follower coarse metal grid, N , sepa-
ring, b. It is suspended rated from the wall by a
from the cover of a “sub- distance of 1.5 cm. (0.59
marine” v e s s e l , B, by inch) and using this grid
Formica hangers, c, . pro- and inner can S as elec-
viding a 4-om. (1.6-mch) trodes. The current
air space between the sub- COLD CALORIMETER WORKING CALORIMETER through coil E was deter-
marine and the calorime- mined by measuring the
ter. The submarine is F I G U R E 3. THERMOELEMENT SYSTEM voltage across a standard
MARCH, 1936 INDUSTRIAL AND ENGINEERING CHEMISTRY 377
one-ohm resistance. A Leeds and Northrup type K-2 potentiome- TIME IN MIWTES
ter was used in all voltage measurements. A period of 600 seconds
was used for heating, and the time was measured with a precision ia
of better than 0.01 per cent by means of a pendulum clock and
automatic switch. R
The temperature change of the solution was measured by means
of a twenty-four-junction thermoelement, W24. This thermo-
element and one of four junctions, to be referred to later, were
calibrated by comparison with a platinum resistance thermometer
which had been calibrated to =+0.02' F. in this laboratorv at
the ice point, steam point, sulfur point, and the transition goint
of Na2S04.10Hz0.
:I 3
In order to maintain all potentiometer readings in the low-
est range where the sensitivity is 0.1 microvolt, the "cold"
junctions of thermoelement Wz4were maintained at a constant
temperature above 32" F. by means of a cascade cold calo-
TIME IN MIWTES
rimeter suggested by White (16). This calorimeter is shown
in Figure 2, and Figure 3 shows diagrammatically the thermo- FIQURE
4. TEMPERATURE
DRIFTS IN A TYPICAL
element system: RUN
Copper block A (Figure 2 ) is contained in a small vacuum flask,
B. Flask B is placed inside a hollow copper block, G, which in agreed within 0.7 per cent with that found by direct measure-
turn is placed inside a wide-mouthed 1-liter vacuum flask, D. ment.
This vacuum flask is incased in a sheet iron vessel, E , which is, in
turn, immersed in a thermostatically controlled bath, F . Each Method of Operation
copper block is provided with a resistance heater, and current
can be passed through these resistances individually or in series. From the foregoing much of the experimental procedure
A sixteen-junction thermoelement, 4 6 , similar t o the twenty-four-
junction element of the adiabatic control, measures the tempera- may be inferred. With a solution of given concentration,
ture difference between the outer block and the bath. A four- several specific heat determinations were made at different tem-
junction thermoelement, C1, with one set of junctions in an ice peratures. The cold calorimeter was first brought to equi-
bath measures the'temperature of the outer block. A twenty- librium a t a temperature about midway in the range to be
four-junction thermoelement, D24, measures the temperature dif-
ference between the inner and outer copper blocks. Wheqmeas- covered. I n order to correct for the heat of stirring and for
uring the temperature of the solution, thermoelement 0 2 4 is con- the small temperature drift of the cold calorimeter, each ex-
nected in series with the main working thermoelement, Wzr, so periment consisted of three periods: a fore period, a heating
that the outer block temperature cancels and the temperature of period, and an afterdrift period. Readings of time and e. m. f.
the inner block becomes the reference for measuring the solution
temperature. In Figures 2 and 3 the thermoelements are shown of thermoelement W24 +
0 2 4 were made a t intervals for at
as single junctions to avoid confusion. The junctions are grouped least 15 minutes before starting the heating current and for at
and inserted in holes in the blocks. In operation, the tempera-
tures of the two blocks were e ualized by means of their resistance
heaters. The temperature o? bath P was adjusted to the same
temperature and maintained at this temperature by means of I. SUMMARY
TABLE OF HEATCAPACITY
RUNS
thermoregulator I and heater K. The bath was stirred vigor-
ously by stirrer L. The temperature of the thermostat was auto- NaOH Temp Capacity
Heat NaOH Temp. Capacity
Heat
matically maintained constant within *O.OIOo F. H is an Teight 70 ' F. B. t. u./lb.f' F . Weight $' & F. B.t. u./lb./OF.
auxiliary cooling coil. . .
At 131" F. and with the thermostat adjusted so that the dif- 4.081 37.62 0.9380 41.48 68.11 0,8126
46.89 0.9416 77.02 0.8143
ference between the two blocks remained zero as indicated by 68.00 0.9476 86.27 0.8148
thermoelement Dz( over a 3-hour period, the actual temperature 86,OO 0.9523 47.71 68.11 0.7857
of the blocks as indicated by thermoelement C4 was observed at 6.920 39.49 0.9113 76.75 0.7840
frequent intervals and found to increase linearly with time at a 44.89 0.9141 86.11 0,7849
55.89 0.9184
rate of 0,0003"F. per minute. The rate of increase was greater 76.95 0.9254 5.78 106.18 0.9406
at higher temperatures. It is possible that chemical action 122,50 0.9428
21.38 42.73 0.8465 140.09 0.9454
caused the drift since the copper blocks were covered with a light 43.12 0.8473 154.27 0 9476
film of oil. No error was introduced in the measurements of 50.36 0.8506 11.90 103.80 0.9027
51.66 0.8520
temperature changes of the solutions by this linear drift, since it 61.77 0.8566 124.20 0.9040
is eliminated in the same manner as the drift due to stirring. 69,04 0.8590 139.23 0.9089
77.31 0.8623 154.67 0.9108
86,36 0.8666 19.89 107.10 0.8776
Solutions 29.71 46.54 0.8286 120,47
140.49
0.8802
0.8840
58.21 0.8344 158.38 0.8857
Thermoprene-lined ( I d ) glass, nickel, or monel metal ves- 37,53 42.21 0.8150
0.8172 29.53 103.78 0.8680
49,73 122,41
sels were used in preparing the solutions. Stock solutions 58.55 0.8208 140.11
0,8633
0.8676
were made by first preparing crystalline NaOH.H1O from 50.88 55.98
68.22
0.7685
0.7670
155.41 0.8688
c. P. stick sodium hydroxide and redissolving the solid to 77.90 0.7661 39.44 103,05 0.8286
93.02 0.7651 105.75 0.8294
produce a saturated solution a t room temperature. The in- 121.96 0.8320
6.80 67,82 0.9230 140.14 0.8337
soluble sodium carbonate was removed by settling. The 7.39 67.69 0.9179 156.11 0.8348
solutions were analyzed gravimetrically with an accuracy of 69.48 0.9184 About 44%
0,9070
81.59 0.8036
8.91 67.91 not ana- 103.48 0.8052
*0.005 per cent by the method suggested by Richards and 17.70 68.47 0.8695 lyaed 122.18 0.8067
Hall (9). 78.87 0.8731 51.15 122.22 0.7606
20.98 68.47 0.8607 140,04 0.7596
77,ll 0,8645 165.77 0.7587
Heat Capacity of Calorimeter 86,OO 0.8683 45 20 142.32 0.7997
31.52 68.32 0.8372 157.37 0.7981
The heat capacity of the calorimeter was determined elec- 77.40 0.8413
0,8451
176.27 0.7971
86.40 48.70 150.67 0.7744
trically, using water, in a manner identical to that used in 37.10 67.95 0,8248 176.32 0.7728
measuring the heat capacities of the solutions. Measure- 77.38 0.8280 191.37 0.7723
ments were made in a range from 40' to 140" F. At 68" F. 40.44 68.29 0.8153
77.02 0.8179
the heat capacity was also calculated from the heat capacities 86.04 0,8198
and weights of the constituent parts, and the value obtained
318 INDUSTRIAL AND ENGINEERING CHEMISTRY VOL. 28, NO. 3

of the calculated evaporation correc-

TABLE11. SPECIFIC HEATSOF SODIVM HYFROXIDE
SOLUTIONS tion was applied to measurements in
IN B. T. u. PER POUND
PER F. which evaporation inside of vessel A
Wei ht was appreciable, although the correc-
Per 8ent ,-- Temperature, OF.:
NaOH 32 40 50 60 80 100 120 140 160 180 200 tion was in any case very small.
0 1.034 1.003 1.001 0.999 0.998 0.997 0.998 0.999 1.000 1.002 1.004
2 0.965 0.967 0.968 0,969 0.972 0.974 0.977 0.978 0.980 0.983 0.986
4
6
0.936 0.940 0.943
0.914 0.920 0.924
0.946
0.928
0.951 0.954
0.933 0.938
0.957
0.941
0.960
0.944
0.962
0.946
0.965
0.948
0.966
0,950 Results of Measurements
8 0.897 0,902 0.907 0,911 0.918 0.923 0.927 0.930 0 932 0.934 0.936
10 0.882 0.888 0.893 0.897 0.905 0.911 0.916 0.918 0.920 0.922 0,923 The results of eighty-one measure-
12 0.894 0.901 0.906 0,909 0.911 0.912 0.913 ments are shown in Table I. Few
14 0.886 0.892 0.897 0.901 0.903 0.903 0.904
16 0,880 0.886 0.891 0.894 0.896 0.897 0.897 measurements were made above 158"F.,
18 0.873 0.880 0.885 0.888 0.890 0.891 0.891
20 0.868 0.875 0.880 0.884 0.886 0.886 0.887 but it was found that above 140" F.
22 0.863 0.870 0.876 0.880 0.882 0,882 0.883 the heat capacities were practically in-
24 0.858 0.866 0.873 0.877 0.879 0.879 0.880
26 0.854 0.863 0.869 0.874 0.875 0.876 0.876 dependent of temperature so that little
28 0.850 0.859 0.866 0.870 0.872 0.872 0.873 uncertainty was involved in extrapo-
30 0.846 0.855 0.862 0.866 0.868 0.869 0.869
32 0.842 0 850 0.857 0.862 0.863 0.864 0.864 lating to higher temperatures. The
.. .. .. . .. .
34 . . 0.823 0.828 0.837 0.845 0.852 0.856 0.857 0.858 0.858 temperature coefficient of the heat ca-
36 0.819 0.824 0.832 0.840 0.845 0.849 0.850 0.851 0.851
38 ... . . . 0.816 0.820 0.827 0.833 0.837 0.841 0.842 0.842 0,843 pacity near 32' F. is of considerable
40 ... . . . 0.812 0.815 0.821 0.826 0.829 0.831 0.832 0.832 0,832
42 ... ...
0.807 0.809 0.813 0.816 0.819 0.819 0.820 0.820 0.820 magnitude and changes rapidly with
44
46
...
...... . . .
... ...
0.802
0.793
0.804 0.806
0.794 0.795
0.807
0.794
0.807
0.794
0.807
0.793
0.806
0.791
0.804
0.789
temperature so that extrapolation to
... ... 32" F. was comparatively unsatisfac-
.. .. ..
48 ... 0.783 0.782 0.781 0.780 0.779 0.777 0,776
50 ... ... ... 0.771 0,769 0.768 0.767 0.765 0.765 0.764 tory. The heat capacities were con-
52 ... .., ... ... 0.758 0.756 0.755 0.754 0.753 0.752 0,752
verted to B. t. u. per pound per " F.,
were cross-plotted on a l a r g e s c a l e ,
and were rounded to even values of
least 20 minutes after stopping the current. On plotting concentration and temperature. The results are given in
these points and extrapolating the nearly parallel lines so ob- Table 11. The conversion factor used was 4.1876 joules per
tained to the middle of the heating period, the difference in gram calorie,
e. m. f. values a t this point gave the true change in e. m. f. due
to the electrical energy input. The trend of the afterdrift Comparison with Previous Work
immediately after stopping the heating current indicated that
the temperature of the metal parts of the calorimeter lagged Figure 5 shows a comparison of the heat capacity values of
perhaps 0.04"to 0.05' F. behind that of the solution. Figure Richards and Gucker (8) a t 68" F. with those of the present
4 shows the e. m. f. us. time curves for a tyDical run in which work a t the same temperature. The agreement is well within
t 6 e h e a t i n g wa s the probable precision of the measurements.
started a t time 37 A similar comparison was made with the specific heats
minutes and measured by Gucker and Schminke (3) a t 77" F. The maxi-
stopped a t time 47 mum difference found was 0.13 per cent a t a concentration of
minutes. The de- 4 per cent.
creasing values of In Figure 6 are plotted the specific heat data of Tucker
e. m. f . with time (1.4) at a mean temperature of 68" F., of Thomsen (IS) a t a
were due to the fact mean temperature of 63.5"F., and those of the present work a t
that the cold junc- 68" F. Thornsen's values would be increased slightly if cor-
tions of thermoele- rected to 68" F., so his work is in good agreement. Tucker's
~~
m e n t W24 4- 0 2 4 data, over a considerable concentration range, lie on a curve
1 1 1 1 I\-, I were at a higher almost parallel with, but considerable lower than, that of the
091
1 I ,'\i temperature t h a n present work. Apparently an almost constant error entered
I
t h o s e in the solu- into his results.
tion. A s m a y be
seen from Figure 4, Literature Cited
the temperature of
the solution in the (1: Bertetti and McCabe, IND.ENG.CHEM.,28, 247 (1936).
region of thermoele- (1A) Daniels, J . Am. Chem. SOC.,38, 1473 (1916).
(2) Derby and Marden, Ibid., 35, 1767 (1913).
ment Wzd decreased (3) Gucker and Schminke, Ibid., 55, 1013 (1933).
slightly i m m e d i- (4) McCabe, Trans. Am. I n s t . Chem. Engrs., 31, 129 (1935).
ately after stopping (5) Merkel, 2. Ver. deut. Ing., 72, 109 (1928).
the heating current. (6) Pratt, J. Franklin Inst., 185, 663 (1918).
(7) Richards and Gucker, J . Am. Chem. SOC., 47, 1876 (1925).
The d i s t a n c e A B (8) Ihjd., 51, 712 (1929).
r e p r e s e n t s a tem- (9) Richards and Hall, Ibid., 51, 707 (1929).
perature difference (10) Ibid., 51, 731 (1929).
of a p p r o x i m a t e l y (11) Richards and Rowe, Ibid., 43, 770 (1921).
(12) Soule, IND.ENG.CHEM.,Anal. E d . , 1, 109 (1929).
0.02" F., and this (13) Thomsen, Ann. Physik, 142, 337 (1811).
temperature change (14) Tucker, Trans. Roy. SOC.(London), A215, 319 (1915).
is accounted for by (15) White, J . Am. Chern. SOC., 36, 2292 (1914).
the equalization of (16) White, "Modern Calorimeter," p. 131, New York, Chemical
Catalog Co., 1928.
0 5 10 15 20 25 30 35 40 45 Y)
temperature after
WElGHT PER C E H l NaOH. the heating current RECEIVED August 3, 1935. Abstracted from a dissertation submitted by
FIGURE 6. COMPARIEON O F SPECIFIC
was turned off. John W. Bertetti i n partial fulfillment of the requirements for the degree of
HEATDATA Eighty per cent doctor of philosophy, University of Michigan.