TR N fl

CENTRAT RESEARCH LIBRARY ORNL-TM-4308

AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA

[ANARAR o

3 yy56 0555980 1

DENSITY AND VISCOSITY OF SEVERAL
MOLTEN FLUORIDE MIXTURES

Stanley Cantor

 

 

 

 

 

 

 

 

 

 

OAK RIDGE NATIONAL LABORATORY '
CENTRAL RESEARCH LIBRARY
| DOCUMENT COLLECTION .
LIBRARY LOAN COPY
DO NOT TRANSFER TO ANOTHER PERSON
If you wish someone else to see this
document, send in name with document
and the library will arrange a loan.

 

 

OAK RIDGE NATINAL LABORATORY

OPERATED BY UNION CARBIDE CORPORATION e« FOR THE U.S. ATOMIC ENERGY COMMISSION

 

 

 

 

 
 

 

This report was prepared as an account of work sponsored by the United
States Government. Neither the United States nor the United States Atomic
Energy Commission, nor any of their employees, nor any of their contractors,
subcontractors, or their employees, makes any warranty, express or implied, or
assumes any legal liability or responsibility for the accuracy, completeness or
usefulness of any information, apparatus, product or process disclosed, or
represents that its use would not infringe privately owned rights.

 

 

 
ORNL-TM-L308

Contract No. W-7L05-eng-26

CHEMICAL TECHNOLOGY DIVISION

DENSITY AND VISCOSITY OF SEVERAL MOLTEN
FLUORIDE MIXTURES

Stanley Cantor

March 1973

OAK RIDGE NATIONAL LABORATORY
Oak Ridge, Tennessee 37830
operated by
UNION CARBIDE CORPORATICN
for the
U.S. ATOMIC ENERGY COMMISSION

OAK RIDGE NATIONAL LAB

AR ARTI

3 4456 0555980 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
i1l
CONTENTS

ABSTRACT .
DENSITY OF MOLTEN SALTS

Experimental
Materials . . . .

Results .

Table 1A. Density of LiF-BeFp-ThF,
(70.11-23.88-6.01 mole %)

Table 1B. Density of LiF-BeF)-ThF,
(70.06-17.96-11.98 mole %)

Table 1C. Density of LiF-BeF-ThFy
(69.98-14.99-15.03 mole %).

Table 1D. Density of LiF-BeF
(66~34 mole %) .

Table 1E. Density of LiF-BeFy-ZrF,
(64.7-30.1-5.2 mole %)

Table 1F. Density of LiF-BeFp-ZrF,-UF,
(64.79-29.96-4.99-0.26 mole %)

Table 1G. Density of NaBF4-NaF (92-8 mole %)

Table 2. Density of KNOg3 . . . . .

Discussion .+ . . .+ & v v v v b e e e e e e e e e e e
Additive Molar Volumes . . . .« . . . .
Table 3. Molar Volumes of Fluoride Mixtures
Expansivity

Room-Temperature Density and Estimated Density Change,
Upon Melting, of MSBR Fuel and Coolant Salts

VISCOSITY .

Introduction and Experimental .
Results and Discussion . . . . . . .

Table 4. Viscosity of NaBF,-NaF
(92~-8 mole %) . « + « + o .
Table 5. Viscosity of LiF-BeF»-ThFy,
(72.7-15.7-11.6 mole %)
Table 6. Viscosity of LiF-BeFy-ThFy
(70.11-23.88-6.01 mole %) .
Table 7. Viscosity at 800°K and 900°K of LiF- BeFZ—ThF4
(OrUFa) P

REFERENCES . . . . . . . .

Page

10
11
12

13
13

14
15

15
19

19
20

21
22
23

25
26
DENSITY AND VISCOSITY OF SEVERAL MOLTEN FLUORIDE MIXTURES
Stanley Cantor
ABSTRACT

Using a dilatometric method, densities were determined for the follow-
ing molten salts:

LiF-BeFy (66-34 mole %)

LiF-BeFy-ThF, (70.1-23.9-6.0, 70-18-12, 70-15-15 mole %)
LiF-BeFy-ZrF, (64.7-30.1-5.2 mole %)

LiF-BeFp-ZrF4-UF, (64.79-29.96~4.99-0.26 mole %)
NaBF;-NaF (92-8 mole %)

KNO3

The last salt was measured to assure the accuracy of the method; the densi-
ties measured for KNOj agreed within 0.15% with critically evaluated
densities obtained by Archimedean methods.

For the fluorides, molar volumes obtained from the density measure-
ments agreed within 27 with volumes calculated from additive contributions
of the components. The expansivities of three LiF-BeF)-ThF4 mixtures were
practically identical, 2.5 x 10~4/°c.

Density-temperature curves from 25-700°C for LiF-BeF2-ThFy (72-16-12
mole %) and for NaBF,-NaF (92-8 mole %) were derived from room-temperature
pycnometric determinations and from estimated expamsivities of the solid
salts. The calculated expansion, upon melting, for the former is 7%, for
the latter 8Z.

Viscosities of three salt mixtures were determined by oscillating-cup
methods:

LiF—BeFZ—ThF4 (72,7-15.7-11.6, 70.1-23.9-6.0 mole %)
NaBF,-NaF (92-8 mole %)

Viscosity measurements were conducted at Mound Laboratory, Miamisburg, Ohio,
using capsules and samples prepared at ORNL. The viscosities of the two
melts composed of LiF, BeFy, and ThF, were analogous to viscosities report-
ed for similar mixtures containing U#4 instead of ThF4.
DENSITY OF MOLTEN SALTS

The objectives of this investigation were: (a) to measure, with high
accuracy, densities and expansivities of several molten fluoride mixtures
that are significant to molten-salt reactors, (b) to derive additive molar
volume contributions which can serve to predict densities in
LiF-BeF)-(Th,U)F, molten solutions, (c) to estimate density changes upon
melting of the fuel-carrier and coolant salts of the molten-~salt breeder

reactor.
Experimental

Apparatus and Procedures: Densities were determined in a nickel dilato-
meter (Figure 1), the details of which have been previously described.l’2
In the apparatus a metal probe detects the changes in liquid level in the
neck of a volume-calibrated metal vessel. The escape of vapor is prevented
by a Teflon stopper, which also permits vertical displacement of the probe.
When the probe contacts the liquid surface, a vacuum-tube voltmeter changes
from an open circuit reading to a detectable resistance. The probe height
is measured to + 0.02 mm with a cathetometer. Through a side arm in the
neck of the vessel, an inert insoluble gas (argon) is introduced to sup-
press bubbles in the melt. By taking measurements at argon pressures of
approximately 5 atm, entrapped gas bubble volumes were reduced to less than

0.1% of the liquid sample.

After completing measurements at elevated tempefatures, the contents
of the vessel were removed and weighed to be certain that weight changes
had been negligible. For any sample measured, weight changes never exceed-
ed 0.05%Z. After each run, the dilatometric vessel was recalibrated at room
temperature with distilled water. The recalibrations indicated that the

nickel vessels sustained permanent expansions of about 0.2%.

Melt temperatures were controlled to + 0.2° by regulating the furnace
with a Leeds and Northrup Speedomax proportional controller. Temperatures
of the melt were determined with Pt-Rh thermocouples previously calibrated
by the National Bureau of Standards; these thermocouples are stated to be

accurate within 0.5° in the temperature range (400-820°C) of measurement.
ORNL-DWG 69-1541

  
  
   
    
 

ROD FOR ATTACHMENT TO RACK

| +— MACHINED LINE

ELECTRICAL LEADS
TO VACUUM-TUBE

VOLTMETER TEFLON STOPPER

SWAGELOK UNION

 

 

TO VACUUM PROBE
OR &_—— /
INERT GAS 4
L THERMOCOUPLE
]
Ve WELL
FURNACE —

 

 

 

 

 

 

£

 

 

CERAMIC
PEDESTAL

 

 

 

 

 

 

 

 

Fig., 1. Dilatometer for Measuring Volume of Molten Salts. The
part of the probe above the Teflon stopper is longer than indicated
in the figure,
Materials

The salt mixtures, LiF-BeF, (66-34 mole %) and LiF-BeFy-ZrF, (64.7-
30.1-5.2 mole %), were supplied by J. H. Shaffer, ORNL, from batches that
had beenASparged Ey Ho-HF gaseous treatment.3 By adding purified, crystal-
line LiF and UF,4 to the latter, we prepared LiF-BeFy-ZrF,-UF, (64.79-
29.96-4.99-0.26 mole 7). In the order given above, the three compositions

corresponded to the MSRE coolant, carrier salt, and fuel salt mixtures.

Mixtures of LiF-BeF)-ThF, (used in both density and viscosity measure-
ments) were constituted from LiF-BeF, (66-34 mole %), crystalline LiF, and
LiF-ThF, mixtures. The LiF-ThF,, retained from a previous density study,4
had been stored in a vacuum desiccator. The mixture, NaBF;-NaF (92-8 mole %),

. e 2
was constituted from the purified components. 0

Analytical-grade KNO3 (J. T. Baker Chemical Co., Phillipsburg, N.J.)
was used as received. Measurements of the salt were carried out primarily
for checking the accuracy of the dilatometric method used in the present

investigation.

Argon gas, used for suppressing bubbles in the melts (see above), was
obtained from Airco, Chester, W. Va. The gas was shown by mass spectro-
graphic analysis to exceed 99.9% in purity. Prior to entry into the
vessel, the gas was passed through molecular sieve to remove traces of

moisture.

All salt loadings were carried out in a glovebox filled with helium.
To insure that the liquid level reached into the neck of the vessel (see
Figure 1), two loadings were usually required; after melting the initial
charge of the salt, the vessel was returned to the glovebox for further

loading.
Results

The densities of seven fluoride mixtures were measured over about
a 200°C temperature range. The data are listed in Tables 1A - 1G; also
included are the least-squares equation and the densities calculated from
the equation. For each melt the plot of density versus temperature was

linear. Data for KNO,'l are given in Table 2,
Table 1A. Density of LiF-Ber---ThF4 (70.11-23,88-6.01 mole %)

 

 

 

Temperature | Density (g/cm3)

(°C) Experimental Calculateda
555.1 2.7406 2.7395
571.9 2.7276 2,7282
580.9 2.7238 2,7222
596.7 2.711l 2.7116
606.2 2.7049 2.7052
621.3 2.6940 2.6951
628.7 2.6901 2,6901
646.8 2.6774 2,6780
655.1 2.671l 2.6724
673.2 2.6606 2.6603
681.2 2.6536 2.6549
707.4 2.6398 2,6372

 

®From the least-squares equation:

o(g/cm’) = 3.1118 - 6.707 x 10"t (°C).
Table 1B. Density of LiF—-Ber—-ThF4 (70.06-17.96-11.98 mole %)

 

 

 

Temperature Density (g/cm>)

(°C) Experimental Calculateda
533.2 3.3942 3.3936
558.1 3.3730 3.3735
561.4 3.3698 3.3709
580.7 3.3551 3.3553
588.8 3.3481 3.3488
603.4 3.3364 3.3370
615.0 3.3278 3.3277
626.5 3.3198 3.3184
640, 2 3.3060 3.3073
649.6 3.3009 3.2997
673.1 3.2837 3.2808
696.9 3.2630 3.2616
721.0 3.2412 3.2422
741.2 3.2237 3.2259

 

a .
From the least-squares equation:

og/cm>) = 3.8236 - 8.064 x 10™*t (°C).
Table 1C. Density of L1F-BeF,~ThF, (69.98 - 14.99 - 15,03 mole %)

 

 

 

Temperature Density (g/cmB)
(°C) Experimental = Calculated?d
543.4 3.6632 3.6634
556.7 3.6493 3.6507
582.9 3.6242 3.6258
608.5 3.6021 3.6014
620.9 3.5897 3.5896
633.7 3.5788 3.5774
646.4 3.5668 3.5653
659.1 3.5541 3.5532
672.6 3.5413 3.5403
698.9 3.5163 3.5153
713.7 3.5000 3.5012
730.2 3.4836 3.4854
749.5 3.4665 3.4671

 

%From the least-squares equation:

4

o(g/em>) = 4.1811 - 9.526 x 10 %t (°C).
Table 1D. Density of LiF-BeF, (66-34 mole %)

 

 

 

Temperature Density (g/cmB)

(°C) Experimental Calculatedad
514.5 2.0292 2.0284
540.5 2.0153 2,0157
564.9 2.0030 2.0038
590.5 1.9915 1.9913
614.6 1.9797 1.9795
616.0 1..9785 1.9788
667.1 - 1.9540 1.9539
719.5 1.9285 1.9283
772.2 1.9027 1.9025
794.7 1.891l 1,8915
820.3 1.8792 1.8790

 

%From the least-squares equation:

o(g/cm) = 2.2797 - 4.884 x 10t (°C).
Table 1E, Density of LiF-Ber—ZrF4 (64,7-30.1-5.2 mole %)

 

 

 

Temperature Density (g/cm3)

(°C) Experimental Calculatedg;_
452,0 2.2780 2.2780
475.8 2.2628 2.2642
501.0 2.2497 2.2497
503.5 2.2481 2,2483
523.4 2.2371 2.2368
530.6 2.2320 2.2326
546,9 2,223, 2.2232
570.8 2.2096 2.2094
594.9 2.1960 2,1955
597.7 2.1940 2,1939
619.0 2.1822 2.1816
622.6 2.1813 2.1796
642,2 2.169, 2.1682
647.5 2.166 2.1652
666.5 2.1547 2.1542
672.4 2.1489 2.1508
698.2 2.1350 2.1359
703.9 2.1314 2,1327

 

%From the least—-squares equation:

p(g/cmB) = 2,5387 - 5.769 x 10-4t (°C).
10

Table 1F. Density of LiF-BeF,.~ZrF —UF4 (64.79-29.96-4.99-0.26 mole %)

 

 

 

2 4
Temperature Density (g/cmB)
(°0) Experimental Calculated®
524.3 2.257¢ 2.,2587
571.1 2.231q 2.2324
617.2 2.2057 2.2064
625.6 2.2054 2.2017
640.7 2.1928 2.1932
664.1 2.1800 2.1801
697.5 2.1626 2.1613
715.8 2.1493 2.1510
761.1 2.125l 2,1256

 

a .
From the least-squares equation:

o(g/em>) = 2.5533 - 5.620 x 10 't (°C).
11

Table 1G. Density of NaBF4—NaF (92-8 mole %)

 

 

 

Temperature Density (g/cm3)

(°C) Experimental Calculatedd
399.5 1.965, 1.9680
423.4 ' 1.950, 1.9511
448,0 1.936, 1.9336
471.9 1.918, 1,9166
494,6 1.901,4 1.9004
495.8 1.900, 1.8996
519.8 1.882, 1.8825
543.4 1.8664 1.8657
567.4 1.8466 1.8487
590.8 . 1.8314 1,8320

 

4¥rom the least-squares equation:

(g/cm3) = 2.2521 - 7,110 x 10 't (°C).
12

Table 2. Density of KNO

 

 

 

3
Temperature L Density (g/cmB)
(°cy Experimental Calculated@
343.6 1.8716 1.8695
360.4 1.8579 1.8571
369.8 1.8503 1.8501
375.2 1.8456 1.8461
384.0 1.839l 1.8395
384.6 1.8380 1.8391
386.4 1.8374 1.8378
389.0 1.8348 1.8358
395.8 1.8307 1.8308
399.5 1.8275 1.8280
403.1 1.8239 1.8253
412.6 1.817l 1.8183
414.8 1.8179 1.8167
416.8 1.8185 1.8152
425.9 l.8068 1.8084
426.0 1.8090 1.8083
437.7 1.8000 1.7996
445, 8 1.7933 1.7936
450.9 l.7894 1.7898
474.0 1.772l 1.7727
499.4 1.7543 1.7538
511.8 1.7443 1.7446
537.9 1.7252 1.7252
560.3 1.7087 1.7086
586.3 1.6893 1.6893
611.9 1.6707 1.6702

 

a .
From the least—-squares eguation:

Q(g/cmB) = 2.1248 - 7.428 x 10*4t (°c).
13

The standard error in density was approximately 0.001 g/cm3, corres-
ponding to about 0.05%. Other sources of error (creep sustained by the
vessel, bubble volume, small amounts of salt condensed on the upper neck
of the vessel) increase the total error to + 0.3%Z. This percentage error
was determined by comparing our results with those of Bloom et §i47
Janz,8 in his critical review, judges the uncertainty in Bloom's results
to be about 0.27%; our results differ from those of Bloom by 0.15%. The

density-temperature equations for KNOg are:

p(g/cmB) = 2.116 - 7.29 x 10_4t (°C) Bloom 9£>§£;7
p(g/cmB) = 2.125 - 7.43 x 10—4t (°C) our results.
Discussion

Additive Molar Volumes

 

The simplest, and often quite successful, way for estimating the
density of solutions is to assume that the volume of a mixture is the sum
of additive contributions of the component compounds. The additive con-
tributions are usually available from density measurements of the com-

ponents; the density - and hence the molar volumes, of LiF,4 ThF ? and

4?
BeF2lO have been reported by the author. At 550 and 700°C, the molar

volumes obtained from these investigations are:

Volume (cm3)

 

 

550°C 700°C
LiF 13.24 13.77
BeF 9 24.0 24.2
ThF, 46.15 47.00

Molar volumes of the three LiF-BeF)-ThF, mixtures and the LiF-BeF)
mixture were calculated from the values above; the calculated and experi-
mental molar volumes are compared in Table 3. Calculated volumes are ap-
proximately one percent greater than experimental values. The good agree-
ment is probably due to the small sizes and low polarizabilities of the

ions which comprise these mixtures,

The concentrations of ZrF4 and UF, in the two mixtures studied were

not large enough to test whether or not their molar volume contributions
Table 3.

Molar2 Volumes of Fluoride Mixtures

Molar Volumes (cm3)

 

 

 

 

Composition (mole fraction, Ni) — 530°¢ = 700°C

LiF BeF, ThF, Exptl. Calcd. piff.®  Exptl.  Calcd. Diff.©

0.7011 0.2388  0.0601 17.47  17.7, 1.89% 18.14 18.2, 0.88%

0.7006 0.1796  0.1198 18.79  19.1, 1.65% 19.49 19.6, 0.56%

0.6998  0.1499  0.1503 19.55  19.8, 1.79% 20. 34 20. 3, -0.15%

0.66 0.34 - 16.46  16.9, 2,67% 17.08 17.3, 1.41%

LiF BeF, ZrF,

0.647 0,301 0.052 17.84  18.1g 1.91% 18.56 18.6,, 0.22%

0.6479  0.2996  0.0499 17.85  18.1, 1.85% 18.54 18.6, 0.81%
+0. 0026UF,,

NaBF4 NaF

0.92  0.08 56.08  56.1p 0.05% | 59.49¢  59.7; 0.37%

 

 

8\ mole of salt mixture is defined: M = J NiMi’ where M is

bCalculated from the equation V = I N;iVi, where V and V{ are, respectively, molar volumes
of the mixture and of component i, both at the same temperature.

Discussion.

c

100 x

 

experimental volume

dExtrapolated.

(Calculated volume minus experimental volume).

 

molar mass, Ni is mole
fraction of component i, Mi is gram—-formula weight of component i.

Values of Vi given in the

71
15

were additive. Nonetheless, the molar volumes at 550 and 700°C of the
mixtures containing these components were calculated using, in addition to

the LiF and BeFy molar volume given above, the following:

- ZYFy: 46 cm3 at 550°C; 48 cm3 at 700°C
UF,: 45.1 cm3 at 550°C; 46.1 cm> at 700°C

The ZrF, volumes were derived (not measured directly) from the densities
of alkali fluorides - ZrF, melts studied by Mellors and Senderoff.ll Molar
volumes for UF4 are extrapolated from densities measured by Kirshenbaum

and Cahill.12

For NaBFAvNaF (92-8 mole %), the observed molar volume would not be
expected to deviate from the additive value, Table 3 shows that volumes
calculated from additive contributions agree within 0.4% with experimental

s

- . » . 2 l
results. The additive contributions 3 are:

NaBFA: 59.35 cm3 at 550°C; 63.20 cm3 at 700°C

NaF : 18.82 cm° at 550°C: 19.62 cm° at 700°C

Exgansivitz

An interesting result, derived from the three mixtures containing
ThFa, is that the expansivity (fractional change of volume with tempera-
ture) did not seem to change with the concentrations of BeF2 and ThFa.
Given that any fuel mixture for a molten~salt breeder reactor will contain
about 70 mole % LiF, then the results suggest that the expansivity will be
very close to 2.5 x 10—4/°C. The actual results were:

-1 3p

Salt Composition Expansivity, o = 53T at 600°C
(mole %) Units are (°C)'l
70.11 LiF, 23.88 BeF,, 6.0l ThF, 2.4g x 10'2
70.06 LiF, 17.96 BeF,, 11.98 ThF, 2.4y x 10
69.98 LiF, 14.99 BeF,, 15.03 ThF, 2.6, x 1074

Room-Temperature Density and Estimated Density Change, Upon Melting, of

 

MSBR Fuel and Coolant Salts

 

This short investigation was conducted in order to provide reactor
designers with a reasonable estimate of the density change, upon melting,

of MSBR fuel and coolant salts.
16

Densities, at room temperature, were determined pycnometrically in a
25—cm3 Kimax '"specific gravity bottle'". The precise volume of the bottle
was determined with distilled water. Cottonseed oil was used as the dis-
placement liquid for the salts; the latter had been prefused and only
relatively large (> 2 mm) crystalline fragments were used in the pycno-

meter. The results obtained were:

LiF—Ber—ThF4 (72-16-12 mole 7%); 3.7887 g/cm3
NaBF, (100 mole %): 2.435. g/cm’

The pycnometric density of NaBF, was 3% less than the x-ray density of

16
2.5075 reported by Brunton.

A density-temperature curve (Fig. 2) for LiF-BeF --ThF4 (72-16~12 mole

%) was constructed on the basis of the following assuiptions: a) the
pycnometrically determined density at 25°C is representative of the bulk
density of the solid salt; b) the volume expansivity of the solid is

1l x 10-4 /°C, an estimate based on the value of this property in other
salts;15 c) the density above the liquidus is reliably predicted from the

additive molar volumes for LiF, BeF,, and ThF4 listed in the first part of

2’
this report. The calculations result in a predicted 7% decrease in dens-
ity over the temperature range of melting (or freezing). Equations and

other details are noted in Figure 2.

Two curves depicting the density—-temperature behavior of MSBR coolant

(92-8 mole % NaBF,-NaF) are given in Figure 3. The solid lines refer to

“"theoretical" or i-ray densities. At 243°C and at 385°C, the dashed and
solid lines coincide over a range of densities. The curves were generated
with the assumptions: (i) at room temperature the molar volumes are addi-
tive, (ii) the density of this solid mixture is 3% less than the x-ray
density (as was observed pycnometrically for pure NaBFA); (iii) the tem-
perature coefficient of density for the solid is a constant, 2.5 x 10_4/°C;
this coefficient corresponds to an expansivity of 1 x 10'—4 /°C; (iv) the

is

x-ray density of the high-temperature form of crystalline NaBF4

2,17 g/cm3 at 243°C (the same as Bredig16 obtained at 265°C).

On the basis of these four assumptions and experimental data for the

liquid, a density decrease of 8% upon melting is possible; however, a
DENSITY (g/cm®)

17

ORNL-DWG 73-2599

 

.\
\\\\\\“‘~59CUL49

\
TN

}\ /souous
—4

= 3. - 1 -

p=3.79-4x10 " (7 25) \’0-\

Bl (ESTIMATED)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PREDICTED
l 7% DECREASE
444 - 500°C IS REGION IN DENSITY
OF SOLID + LIQUID —
7Y ) I
{
LIQUIDUS 1 o,
. o
p=3.665-5.91x10 * ¢
(IESTIMATlED) ‘
0 100 200 300 400 500 600 70O

TEMPERATURE (°C)

Fig. 2. Density of LiF-Bel,-ThF) (72-16-12 mole %).
DENSITY (g/cm3)

2.6

2.5

2.4

2.3

2.2

2.4

2.0

1.9

1.8

1.7

18

ORNL-DWG 73-2600

 

L

|

!

! |

=2.517-2.5x10"% (+- 25)

SOLip-1 [

 

=

 

SOLID TRANSITION
TEMPERATURE (243°C)

 

 

N
N4 p=2.44-2.5x10"

4 (t-25)

 

 

 

 

 

 

 

 

- 3S00/p ;————MELTING POINT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

| ~J (385°C)
1
p=2.085_2.5x1o—y' T | "THEORETICAL" 8%
(7-243) DECREASE ON MELTING
| / L
MORE PROBABLE 5% N
DECREASE ON MELTING
{y
o%
p=12.252-7M x 10~ % ¢
\\
100 200 300 400 500 600 700

Fig,

12,-.

TEMPERATURE (°C)

Density of NaBF) -Na¥ (92-8 mole %),
19

decrease of about 5% is more likely. It should be noted that this salt
undergoes a rather large density change in the solid at 243°C. The pre-
dicted density decrease at this temperature 1s about 12.7Z%.

VISCOSITY

Introduction and Experimental

 

Viscosity is an important physical property in assessing the heat
transfer performance and fluid dynamics of reactor liquids. In these
regards, information on fluoroborates was of special interest. Viscosity
measurements on molten fluoroborates are relatively difficult because of

their high volatility.

For volatile liquids at elevated temperatures, accurate measurement
of low viscosities (<10 centipoises) are conveniently determined by oscil-
lating-cup viscometry. In this method, a cylindrical vessel, which
encapsulates the sample, is caused to execute torsional oscillations.

The rate at which the amplitude of the oscillations is damped depends on
the viscous drag of the liquid upon the walls of the container. The vis-
cosity is determined through the basic equations of fluid dynamics from:
the damping rate, the period of oscillation, the dimensions of the appara-

tus, and the mass and density of the liquid.

Over a period of several years, L. J. Wittenberg and his co-workers at
Mound Laboratory, Miamisburg, Ohio (an AEC-owned facility), have gained
much experience in measuring molten materials, mainly liquid metals,17 via
oscillating-cup viscometry. Because it was both faster and less expensive
for Mound rather than ORNL to obtain accurate viscosities of fluoroborates
and other fluorides of reactor interest, a purchase order for Mound's
services was obtained. Viscometric measurements and treatment of the data
were performed by L. J. Wittenberg and R. Dewitt of Mound Laboratory.
Preparation of samples, fabrication of capsules and supplementary inter-
pretation of the data were done at ORNL under the supervision of the

author.
20

The viscosities of five salt melts were determined. Two of these,

single-component melts of NaBF4 and KBF,, have been reported and discussed

4?
in another publication.18 The other three, the subjects of this report,
are: NaBF4—NaF: 92-8 mole %
LiF—Ber—ThFé: 72.7-15.7-11.6 mole % and 70.1-23.9-6.0 mole %).
Wittenberg has published details of the general techniqile17 and
methods for treating the data.19 The specific apparatus used for this in-

vestigation is described elsewhere.

Capsules, machined out of nickel stock, consisted of a cap and a
cylindrical cup, the latter with approximate dimensions: 1.75 cm I.D.,
1.85 cm 0.D., and 7.5 cm long. After the dimensions were accurately
measured, the capsules were charged with salt equivalent to about 15 cm3
in the expected temperature range of measurements. Weighing and charging
of samples were carried out in a glovebox. After these operations, the
glovebox was evacuated at V30 Y for 20 hours. After flushing the box
twice with helium (purified by passing through a charcoal trap maintained
at liquid nitrogen temperature), the cap was fuse-welded to the cup by
means of an argon arc torch. During welding, the capsule was kept in a
copper block whose purpose was to absorb most of the heat generated at the
weld. The efficiency of heat removal was indicated by a small piece of
masking tape attached to the cup about 2.5 cm below the weld-work; the
tape did not appear charred or altered by the welding operations. Success
in the heat removal was confirmed by the negligible losses in capsule

weights taken after welding.

Results and Discussion

 

The viscosities of the three salt mixtures are listed and compared
with least-squares values in Tables 4, 5, and 6. The temperature of the
viscosity determination was measured with a chromel-alumel thermocouple
positioned near the capsule but not touching it. At each temperature, at
least two, and usually three sets of amplitude and period measurements
were taken; hence, there is more than one experimental viscosity entry for
each temperature in Tables 4, 5, and 6. The data and least-squares fit

are plotted in Figure 4.
21

Table 4. Viscosity of NaBF4—NaF

(92 - 8 mole %)

Least squares fit: n(cP) = 0.0877 exp (2240/T(°K))

 

T(°C) Experimental n(cP) Calculated n(cP)

409 2.18, 2.15 2.34
411 2,15, 2.20 2.32
418 2,29, 2.29, 2.30 2.24
425 2,05, 2.05, 2,05 2,17
436 2,02, 2.00 2.06
465 1,89, 1.91, 1.86 1.82
491 1.59, 1.59, 1,59 1.64
505 1.45, 1.43, 1.47 1.56
521 1.45, 1.46 1.47
532 1.31, 1,30, 1.31 1.42

408 2,50, 2.38, 2.46 2.35

417 2.27, 2.35, 2.33 2,25
450 2,03, 2.04, 1.96 1.94
474 1,91, 1.96, 2.08 1,76
505 1.77 1.56

537 1.47, 1.48, 1.44 1.39

 
22

Table 5. Viscosity of LiF-BeF,-ThF

2 4
(72.7-15.7-11.6 mole %)

Least-squares fit: n(cP) = 0.1094 exp (4092/T(°K))

 

 

T(°C) Experimental n(cP) Calculated n(cP)
553 14,1, 14.3, 14.1. 15.5
582 12.4, 13.4, 13.0 13.1
613 11.4, 11.2, 11.1 11.1
638 9.74, 9.56, 9.47 9.76
622 11.5, 11.5, 11.4 10.6
588 13.4, 13.5, 13.4 12.7
555 15.5, 16.5, 16.1 15.3
572 13.45, 13.3, 14.15 13.9
649 9.21, 9.79, 9.22 9.25
673 7.74, 7.75, 7.74 8.27

 
23

Table 6. Viscosity of LiF—-oBer-ThF4

(70.11-23,88-6,01 mole %)

Least-squares fit: Nn(cP) = 0.06602 exp (4380/T(°K))

 

 

T (°C) Experimental n(cP) Calculated n(cP)
653 7.30, 7.06, 7.06 7.47

547 14,1, 13.9, 14,15 13.8

598 9.87, 9.87, 9.88 10.1

633 8.92, 8.97, 8.81 8.30

526 15,39, 16.53, 16.05 15.8

567 12.35, 12.56, 12.35 12,1

579 11.06, 10.92, 10.91 11.3

603 9.69, 10.19, 9.96 9.79

557 12.59, 12,69, 11,85 12.9

 
VISCOSITY (centipoises)

Fig. kL.

20

10

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

  

 

 

 

 

 

 

 

 

 

ORNL-DWG 73-2601
TEMPERATURE (°C)
650 600 550 500 450 425 400
et | U
LiF - BeFp - ThFy
(72.7-15.7-11.6 mole %)
. |
@
vt
3
. A
__-4 = LiF-BeFy-ThFy ]
* 5 (70.14-23.88-6.01 mole %)
“_fl_m | ]
. I
T NOBF4
—""’
e
— : ——
-1
9
] NaBF, - NoF (92-8 mole %)

 

 

 

 

 

 

1
1.05

1.15

1.25
1OOO/T(°K)

1.35

Vigcosities of Three Fluoride Mixtures.
the bottom refer to viscosity of Na.BFLF—NaF (92-8 mole %).

1.45

All the points at
25

The viscosity of NaBF

4
NaBF4.18 Evidently the presence of 8 mole % NaF leads to a somewhat lesser

-NaF (92-8 mole %) is very much like that of

viscosity (see Fig. 4).

The data for the two ternary melts show that the mixture with the
greater viscosity contains the lesser concentration of BeF2 and the greater
concentration of ThFa. Qualitatively, this beggvior was predicted from the
viscosity measured in mixtures of LiF-BeF,-UF, =~ : “"for LiF concentrations
of 60 mole % or greater, substitution of UF4 (or ThF4) for BeF, (at const.
temperature) causes an increase in viscosity."21 Indeed, as Table 7

reveals, the UF4—containing mixtures serve as an excellent basis for quan-

titatively predicting viscosities in analogous ThFa-containing melts.

Table 7. Viscosity at 800°K and 900°K of LiF—Ber—ThF4 (or UF4)

 

 

 

Composition Viscosity (cP)
(mole %) 800°K 900°K
LiF-—Ber—ThF4
72.7-15.7-11.6 18.2 10.3
LiF-BeF,-UF,°
2 74 a a
70-18-12 18.9 lO.4
L:'LF—Ber--ThF4
70.11-23.88-6.01 15.8 §.58
. | a
L1F—BeF2—UF4 . .
70-24-6 _ 18.5 lO.l

 

aSee Reference 20.
10.

11.

12.

13.

26

REFERENCES

S. Cantor, ''Metal Dilatometer for Determining Density and Expansivity
of Volatile Liquids at Elevated Temperature,'" Rev., Sci. Instr. 40, 967
(1969). —_

 

S. Cantor, D. P. McDermott, and L. 0. Gilpatrick, '"Volumetric
Properties of Molten and Crystalline Alkali Fluoroborates,'" J. Chem.
Phys. 52, 4600 (1970).

J. H. Shaffer, Preparation and Handling of Salt Mixtures for the
Molten Salt Reactor Experiment, ORNL-4616 (January 1971).

 

 

D. G. Hill, S. Cantor, and W. T. Ward, '"Molar Volumes in the LiF-
ThF4 System," J. Inorg. Nucl. Chem. 29, 241 (1967).

 

S. Cantor and T. S. Carlton, "Freezing Point Depressions in Sodium
Fluoride. II. Effect of Tetravalent Fluorides," J. Phys. Chem. 66,
2711 (1962). T

S. Cantor, "Freezing Point Depressions in Sodium Fluoride. Effect
of Alkaline Earth Fluorides,'" J. Phys. Chem. 65, 2208 (1961).

H. Bloom, I. W. Knaggs, J. J. Molloy, and D. Welch, "Molten Salt
Mixtures Part I. Electrical Conductivities, Activation Energies of
Ionic Migration and Molar Volumes of Molten Binary Halide Mixtures,"
Trang. Faraday Soc. 49, 1459 (1953).

 

G. J. Janz et al., "Molten Salts: Volume 1, Electrical Conductance,
Density, and Viscosity Data," National Bureau of Standards, Natl.
Std. Ref. Data Ser. No. NSRDS-NBS-15 (U.S. Government Printing

Of fice, Washington, D. C., 1968).

 

 

S. Cantor, D. G. Hill, and W. T. Ward, "Density of Molten ThF,,
Increase of Density on Melting," Inorg. Nucl. Chem. Letters 2,
15 (1966). o

 

S. Cantor, W. T. Ward, and C. T. Moynihan, "Viscosity and Density
in Molten BeF»-LiF Solutions," J. Chem. Phys. 50, 2874 (1969).

G. W. Mellors and S. Senderoff, '"The Density and Surface Tension of
Molten Fluorides,' pp. 578-593 in Electrochemistry, Proceedings of
the lst Australian Conference, February 1963, Pergamon Press (1964).

 

 

A. D. Kirshenbaum and J. A. Cahill, "The Density of Molten Thorium
and Uranium Tetrafluorides," J. Inorg. Nucl. Chem..ii, 65 (1961).

 

J. D. Edwards, C. S. Taylor, L. A. Cosgrove, and A. S. Russell,
"Electrical Conductivity and Density of Molten Cryolite with
Additives," J. Electrochem. Soc. 100, 508 (1953).
14.

15.

16.

17.

18.

19.

20.

21,

27
G. D. Brunton, "Refinement of the Structure of NaBF,," Acta Cryst.
B-24, 1703 (1968).

American Institute of Physics Handbook, Second Edition, p. 4:72,
McGraw-Hil1 Company, New York (1963).

 

M. A. Bredig, Chemistry Division Annual Progr. Rept. May 20, 1971,
ORNL-4706, p. 155.

 

L. J. Wittenberg and D. Ofte, "Viscometry and Densitometry - A.
Viscosity of Liquid Metals,'" Physicochemical YMeasurements in Metals
Research, Part 2, R. A. Rapp (ed.), Interscience Publishers,

New York, 1970, pp. 193-217.

 

 

R. Dewitt, L. J. Wittenberg, and S. Cantor, 'Viscosity of Molten
NaCl, NaBF,, and KBF4," accepted for publication in Physics and
Chemistry of Liquids (1973).

 

L. J. Wittenberg, D. Ofte, and C. F. Curtiss, "Fluid Flow of Liquid
Plutonium Alloys in an Oscillating-Cup Viscosimeter," J. Chem. Phys.
48, 3253 (1968).

B. C. Blanke et al., Density and Viscosity of Fused Mixtures of
Lithium, Beryllium, and Uranium Fluorides, Mound Laboratory Report
MLM-1086 (December 1956).

 

 

S. Cantor et al., Physical Properties of Molten-Salt Reactor Fuel,
Coolant, and Flush Salts, ORNL-TM-2316 (August 1968), p. 9.

 
o o2 )

03
A
1

el o
= OO e o)

72,
73.
7k,
75.
76.
77.

78-79.
go.

31.
82.

L 29

INTERNAL DISTRIBUTION

Central Research Library Ly, A. G. Grindell
ORNL -~ Y-12 Technical Library 50. P. N. Haubenreich
Document Reference Section 51. W. R. Huntley
Laboratory Records Department 52, R. B, Lindauer
Laboratory Records Department-RC 53. H. k. McCoy
ORNL, Patent Office 54. A, P. Malinauskas
C. F. Bages 55. L. E. McNeese
C. E. Bamberger 56. A. S. Meyer

5. K. Beall 57. A. M. Perry

M. R. Bennett 58. J. D. Redman

F. G. Bohlmann 59. M. W. Rosenthal
R. B. Briggs 60. H. C. Savage

K. B. Brown 61. C. D. Scott

S. Cantor 62. Dunlap Scott

I. L. Compere 63. J. H. Shaffer
J. L. Crowley 64, F. J. Smith

F. L. Culler 65. G. P. Smith

J. M. Dale 66. R. A. Strehlow
J. H. DeVan 67. D. B. Trauger
J. R. Engel 68, A. M. Weinberg
D. E. Ferguson 69. J. R. Weir

L. M. Ferris 70. J. C. White

W. R. Grimes 7L. R. G. Wymer

EXTERNAL DISTRIBUTION

A, R. DeGrazia, Division of Reactor Research and Development,
USAEC, Washington, D. C. 20545

Norton Haberman, Division of Reactor Research and Development,
USAEC, Washington, D. C. 20545

F. Dee Stevenson, Division of Physical Research, USAEC,
Washington, D. C. 20545

D. F. Cope, USAEC, RDT Site Representative, ORNL, P.O. Box X,

Qak Ridge, Tenn. 37830

Kermit Laughon, USAEC, RDT Site Representative, ORNL, P.O. Box X,
Oak Ridge, Tenn. 37830

Research and Technical Support Division, USAEC, Oak Ridge
Operations, P.0. Box E, Oak Ridge, Tenn. 37830

Technical Information Center, P.0O. Box 62, Oak Ridge, Tenn. 37830
J. A, Acclarri, Continental 01l Company, Ponca City, Cklahoma
7h601

D. R. de Boisblanc, Ebasco Services, Inc., 2 Rector St., New York,
N. Y. 10006

J. W. Koger, Union Carbide Nuclear Corporation, Y-1zZ Plant,

Oak Ridge, Tenn. 37830