Electronic Conversion for Graphic Recording with the Chevenard Photographically Recording Thermobalance SAUL GORDON and CLEMENT CAMPBELL Pyrotechnics Chemical Research Laboratory, Picatinny Arsenal, Dover,
T h e postwar commercial availability of the Chevenard a u t o m a t i c photographically recording thermobalance has given a g r e a t i m p e t u s to the heretofore inadequately exploited thermoanalytical t e c h n i q u e of thermogravimetry. Because of the inconveniences of photographic recording techniques, a m e t h o d has been developed for obtaining a pen-and-ink recording of the thermogravimetrie curve$ w i t h a potentiometrictype recorder. This is accomplished by means of a linear variable differential transformer used a s the transducer, a n d does n o t require modification of the balance.
N. 1.
elcctrical signals has to be aecuriltely linear, lightweight, frictionless in operation, sensitive to very snmll deflections, and relatively inexpensive, and involve only a minimum modification of the balance. The device finally selected was a linear variable differential transformer, used in conjunction with a demodulator, which satisfactorily met all these requirements. Peterson (8) has independently reported the application of this transducer to t,he conversion of any analytioal halsnce t,a an aut.amntic recording microbalance. s
S.
INCE the introduction of thermoanalytical techniques utiliaing B thermahelance, reported by Honde. (7) in 1915, several manually operated instruments have been used in studies of ma, terials suoh as clays, minerals, metallurgieel specimens, and many inorganio and organic compounds. In 1935, Dubois' ( S ) reported the first automatic photographically recording thermohalance; but it was not until the post-World War I1 years that Chevenard ( 2 ) developed' the first commercially available recording thermohalance. With this apparatus Duval and others critically evaluated a multitude of preoipitates which have been proposed for, and applied to, gravimetric analysis. These investigations ( 4 ) have demonst.rated the value of this formerly neglected technique as a research and 2tntLlytical tool for the study of hixh temperature reactions and their kinetics.
i' L
"
Figure 2.
Diagram of Chevenard themobahnee B.
c.
-./
Balanoe beam counterpoise
IM. Mirror 0. Oil pots R. Support rod,
Figure 1. Linear variable differential transformer
S.
The Chevenard thermobslmee ( 5 )makes use of an automatic photographic technique for continuously recording the weight of the system under investigation. ( A mechanical recording system has recently been made svailitlile as optional equipment for this apparatus.) However, photographic recording methods are inherently limited and involve inconveniences such as the need for a dark room in which the camera can be loaded and the photographic paper processed; the working space required for the photographic housing and camem; the inability t o ohserve tho curves during the coume of the reaction under study; unavailability of the record until the photographic prooessing and drying are completed; B lack of reference grid lines on the record; thc lengthy setup time required for each determination; the limibiition of camera drum speeds; and the time-consuming procedure involved in the occasional realignment of the light source, balancc-heam mirror, and camera drum. Therefore, an electronic conversion of the balance was considered in order to provide direct pen-and-ink recording on potentiometric recorders. A suitable transducer to convert balance-heam displacements into
Sample crucible
Figure 3. Chovenard thermobalanoe, t e m p e r a t u r e controlling a n d recording e q u i p m e n t , photographic a n d electronic recording a p p a r a t u s 124
V O L U M E 28, N O . 1, J A N U A R Y 1 9 5 6
125
Table I. ThermogravirnetrieAnalysis of Calcium Nitrate Tetrahydrate Assumed % LOSS Product Ca(N01),.2Ht0 Ca(N0ah Ca0 CaO (dry basis)
Theoretical
Obeerved
15.2
14.4
30.5
30.3
76.2
76.3
65.9
Difference -0.8 -0.2
+O.l
66.0
+O.l
DESCRIFTION OF APPARATUS
Transducer. A linear variable differential transformer is 5 transducer that generates an alternatinE current signal which is ~. L ~ - - - ~ ~ - ~ "
coil, two seoondary'coils, and a n arkaturk & magnetic material. The primary coil, PI, Pz,is energized from a suitable sinusoidal source; the two secondary coils, SL, Sa, are connected so that their output voltaxes are 180' out of phase; and the armature is located 80 that it can alter the relative flux distribution which exists between the primary coil and the two secondary coils. Motion of the armature toward secondary coil, SI,results in an ~~~~~
increased output of one phase, and motion towards S, produces an increased output 180" out of phase. If S, and S2 are identical and the armature is centrally located so that each receives s n equal amount of flux, the voltages induced in these secondary coils will be equal and out of phase and a theoretical output of zero will result. This condition represents the null, or balance point, of the differential transformer. The transformer with armature shown in Figure 1 is the Class 6206-A Atcotran differential transformer with a linear range of f0.5 inch and an accuracy within a t least =kO.S%. The armature used on the thermobalance w u purchased without a shaft and then mounted on a silica tube sliqhtly longer than the armature. The dnmaduIator used to rectify the alternating current signals generated by the transformer is a vibrator-type phasediscriminating eonverter, Atcotran Type 6101-C. Balance. The balance (Figure 2) consists of a wire-supported beam, B, to one end of which is attached a vertical rod, R, su porting the sample, 8,and to the other end a mirror, M . Tl$ mirror is used to reflect alight beam, LML', onto aphotographic paper wrapped around a synchronous motor-driven drum. The light beam deflections are linearly proportional to the change in weight of the sample, AM, and thus a curve is obtained of the changes in weight as a function of time. Oil pots, 0, are used to minimize oscillations of the bdance beam.
TEMPERATURE,
'C.
Figure 6. Thcrmogravirnetrioanalysis of calcium nitrate tetrahydrate (0.2228 gram), 15"/min.
>lose-up of thermobalanee showing electronic conversion
Figure 3 is a picture of the Chevenard thermohalanee with both the photographic and electronic recording units, and equipment for programming and recording the temperature of the furnaoe. h a l s__-_ n m I ~ Rrnorlifinrl fnr olmtmnir "__Figure 4 is a close-up viLW nf the recording, wherein the armature of the differential transformer is suspended from the balance beam by a thread so that it hangs freely within thi9 transformer coil. I_
.
_l_l
___
i__ ~________
* , .
design of the balance causes nonlinearities in the beam displacements when the armature is added. This was compensated for hy turning down the counterpoise on a lathe and bringing i t up to the desired weight hy wrapping it with nickel hire. The transformer coil is supported and vertically positioned by a wooden clamp mounted on a rack-and-pinion mechanism from a micrometer slide comparator. The nonmetallic clamp is used to ensure lineas operation of the transformer. The demodulator, which is operated from a 110-volt, conshnt voltaze regulator, supplies t,he 6-volt, 60-cycle signal required for the primary coil and feeds the rectified direot current secondary voltage into the potentiometric strip chart recorder. The recorder used is a Leeds & Sorthrup Speedomax with adjustable zero and adjustable range.
tal weight changes as recorded with electmnically converted thermobalanee
OPERATION OF BALANCE
..~-.. ..
With the sample in its container positioned on the support rod, and a fractional weight equal to one half the desired weight range
ANALYTICAL CHEMISTRY
126 placed on the calibration platform, the beam is balanced in a horizontal position by means of the counterpoise. This corresponds to the mid-point of the deflection for the full scale change in weight over this range. The coil is positioned with the rack and pinion, so that the armature is a t the null point and a zero voltage is obtained. The recorder is then adjusted so that the voltage indicated is equivalent to the mid-point of the full scale change in weight to be measured. In this way the most linear portions of the balance beam and the differential transformer are used for measuring and recording the weight changes. The calibration consists of placing another equal fractional weight on the platform to obtain the voltage output equivalent to zero weight loss, and then removing both weights to obtain the recorder point corresponding t o a full scale loss in weight. Replacing the fractional weights restores the balance and the recorder pen to the position of zero loss in weight, and the balance is now ready for the determination which is conducted in the normal manner. For reactions involving a gain in weight, the same calibration procedure is followed using a reversed polarity of the transducer output. If recorders with adjustable zero and adjustable range are not available, conventional electronic potentiometric instruments with a range of about 10 mv. can be used (6) with a battery and variable limiting resistor in series m-ith the differential transformer to provide a bucking voltage for obtaining adjustable zero positioning of the pen. A precision potentiometer may be used as an eyternal voltage divider for variable range adjustment. The range of weight changes that can be linearly recorded is eesentially the 400 mg. that can be obtained photographically, and the accuracy over a range of 200 mg. appears to be that involved in reading the record-Le., approximately 0.25%. The
stability, using a source of constant line voltage, is 0.25% over’a period of 8 hours; and the reproducibility over a range of 200 mg. is 0.5 mg. or 0.25qib of the scale, whichever is greater, as indicated by the stepwise addition and removal of fractional weights shown in Figure 5 . The response time is that of the balance-namely, 2 seconds for a 200-mg. change in weight. A typical thermogravimetric curve obtained with the electronically converted Chevenard thermobalance, at a heating rate of 15’ C. per minute, is illustrated by calcium nitrate tetrahydrate shown in Figure 6. The calculated and observed changes in weight are summarized in Table I. LITERATURE CITED (1) Automatic Temperature Control Co., Inc., Philadelphia, Pa.,
Bull. R-31,Eng. Data Sheet No. 73. (2) Chevenard, P., WachB, X., De La Tullaye, R., Bull. soc. chim., (5) 10, 41 (1944). (3) Dubois, P., thesis, University of Paris, No. 2428, June 26, 1935. (4) Duval, C., “Inorganic Thermogravimetric Analysis,” Else\-ier Publishing Co., Amsterdam, 1953.
( 5 ) Ferner Co., Inc., R. Y . (American Agents for Chevenard therino-
balance), hIalden 48, Mass., Brochure. (6) Gordon, S.,and Campbell, C., Bull. Am. Ceram. Soc., in press. (7) Honda, K., Science Repts. Tdhoku Imp. Univ. 4, 97 (1915). (8) Peterson, A, Instruments and Automation 28, 1104-6 (1955). RECEIVED for review July 1, 1955. Accepted September 16, 1955. Division of Analytical Chemistry, 127th meeting ACS, Cincinnati, Ohio, March 1955.
Unsaturation Determination by Acid-Catal yzed Bromination ROBERT E. BYRNE, J R . ~and , JAMES B. JOHNSON Chemical a n d Physical M e t h o d s Laboratory, Carbide a n d Carbon Chemicals Co., Division o f Union Carbide and Carbon Corp., South Charleston, W. Va.
In an attempt to develop a bromination procedure which would be more widely applicable to the determination of unsaturation in organic compounds, the original method of Kaufman has been investigated and modified. The sample is reacted with a solution of bromine and excess sodium bromide in methanol-water medium containing a small amount of hydrochloric acid catalyst. The excess bromine is determined by ronversion to iodine and titration with sodium thiosulfate. The reagent reacts quantitatively and rapidly with a variety of unsaturated compounds. Errors due to substitution are minimized and in most cases results are unaffected by extended reaction time. Correlation between reactivity of the unsaturated compounds and molecular structure is discussed.
H
AkLOGENATIOXprocedures used in these laboratories for the determination of unsaturation include the Wijs (8) method using iodine chloride, the Hanus (2) method employing iodine bromide, bromine in carbon tetrachloride with a mixed catalyst as described by Braae (f), and a modified Francis ( 4 ) method using an acidified solution of potassium bromate-bromide. These reagents are limited in their usefulness because of inaccuracies and unpredictable side reactions. Their use in most cases requires standardized procedures for specific concentrations of the unsaturated compound.
‘ Preqent address, Bakelite Co., Round B r o o k , N. J
Kaufman (3, 6) reported satisfactory results using a solution of bromine in methanol saturated with sodium bromide. It has been stated that bromine added to a solution of sodium bromide forms the probable complex Brl
+ XaBr e KaBr.Bi-2
which is a mild brominating reagent exhibiting little tendency towards substitution. It reacts with unsaturates according to the equation: R-CH=CH-R
+ NaBr.Br,
+
R-CHBrCHBr-R
+ KaBr
Investigation in this laboratory ( 5 ) has shown that the solubility of sodium bromide in methanol is insufficient to provide the excess bromide ion necessary to complex the bromine completely. As pointed out by Uhrig and Levin (?), without excess sodium bromide, use of the reagent is subject to serious error because of substitution. Addition of saturated aqueous sodium bromide to the reaction mixture prior to the introduction of the sample was tried and this modification has been used for some time. Substitution error was greatly reduced in many instances. However, in those cases where scrupulous care was not observed in the addition of the aqueous sodium bromide solution, reproducibility \?as noted to be adversely affected. Examination has shown that this lack of reproducibility may be caused by small variations in sodium bromide concentration or the quantity of aqueous sodium bromide solution used. A more uniform reaction mixture less subject to poor reproducibility was obtained by