Falling Film Molecular StillR. G. Nester Citation: Review of Scientific Instruments 31, 1002 (1960); doi: 10.1063/1.1717099 View online: http://dx.doi.org/10.1063/1.1717099 View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/31/9?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Suppression of thermocapillary instability in a falling film Phys. Fluids 18, 078106 (2006); 10.1063/1.2234798 Still more on Coriolis myths and draining bathtubs—films and video tapes Am. J. Phys. 62, 1063 (1994); 10.1119/1.17660 Onset of nonlinear waves on falling films Phys. Fluids A 1, 1314 (1989); 10.1063/1.857360 Molecular fractionation and jet stills J. Vac. Sci. Technol. 14, 723 (1977); 10.1116/1.569190 Improved Molecular Still and Sublimer for Vacuum Operation Rev. Sci. Instrum. 33, 388 (1962); 10.1063/1.1717858

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

128.114.34.22 On: Sat, 22 Nov 2014 15:27:30

1002 NOTES

is a measure of the apparent line width or the "percent energy resolution," generally used in describing the per­formance of scintillation spectrometers. It has also been found that this relative slope changes in a regular way as a function of the incident gamma-ray energy, which is in agreement with the results of studies on variation in resolution with energy.3 Intermittent monitoring, or sta­bility runs, with two or three common sources like Hg203,

CsI37, and Zn65 thus yield best-fit straight-line probability plots of the data and can serve as references to the relia­bility of the electronic gear.

In making such plots it is necessary to use some judge­ment about what the total number of counts under the photopeak is. Our practice has been to fold carefully upon itself· the differential plot of the photopeak on a linear scale, requiring highest symmetry in the region near the maximum of the peak, and then to use twice the sum of the counts on the high energy side of this fold. 4 This is shown on Fig. 1.

The ultimate straight line on the probability plot is chosen as the best fit for the original information, ignoring the points where skewness appears due to scattering and contributions from the Compton distribution from the crystal. Once the straight line is chosen it is easy to de-

w ...J ~ z w u a:: w a.

0.01.--------------------------------,

0.2 RESOLUTION = I~~:;~ = 7.60 ±.oa %

2.0

10.0

30.0

50.0

70.0

90.0

98.0

99.8

143.85 ±.05

POSITION OF SYMMETRY 138.58 ±.05

99.99L-~~~~~~~...J~~~~~~~~~~~~~~~~~

130 134 138 142 146 150 154

PULSE HEIGHT (ARBITRARY UNITS 1

FIG. 2. Probability plot of CS137 photopeak in Fig. 1. A shows the percentile spread and position of the full width at half-maximum. B shows the skewness introduced on the low pulse-height edge due to scattering and contributions from the Compton distribution from the crystal.

termine the two pulse-height positions for the percentiles (11.95 and 88.05) that determine the width at half­maximum and also the position' for the 50th percentile (the point of symmetry). These are shown on Fig. 2.

In examining the usual linear plot of the differential spectrum, decisions may be unduly influenced by the data points in the neighborhood of the half-maximum counting rate. On the other hand, in the probability plot, the best­fit line uses the total integrated information and hence tends to be less arbitrary. Further uses of the probability plot in spectrometer data study, besides those mentioned, may be found. Hald5 points out some of the possibilities for very general data; specifically in gamma-ray spec­troscopy an added advantage is a priori knowledge of the relative slopes in the plots.

* Supported in part by the joint program of the Office of Naval Research and the U. S. Atomic Energy Commission.

t NSF Science Faculty Fellow 1959/60. On leave from Chico State College, Chico, California.

1 Ernst Breitenberger, Progress in Nuclear Physics (Pergamon Press, London, 1955), vol. 4, p. 56. .

2 Keuffel and Esser No. 359-23, for example. Probability paper has the usual linear scale on its abscissa axis, and a special ruling on its ordinate axis. Plotting the probability integral

y=[I/<T(27r)lJ.f .. exp[ - (x-xo)2/2~]dx,

results in a straight line with a slope = l/er. aN. H. Lazar, IRE Trans. on Nuclear Sci. NS-5, 138 (1958). 4 A poor choice of folding position can be recognized as such be­

cause it results in a distortion from a straight-line plot. 6 A. Hald, Statistical Theory with Engineering Applications (John

Wiley & Sons, Inc., New York, 1952).

Falling Film Molecular Sti11*

R. G. NESTER

E.I. du Pont de Nemours and Company, Wilmington 98, Delaware

(Received May 25, 1960; and in final form, June 27, 1960)

A HIGHLY efficient molecular still of the falling film type (Fig. 1) that combines for the first time a

number of unusual features, some of them novel, is de­scribed. The still is particularly effective for the distilla­tion of high molecular weight heat-sensitive materials and can be operated readily at pressures of 10-5 mm Hg. As a result of its construction, the still is inexpensive to make and easy to clean.

The effectiveness of the still results primarily from the use of a rotating spring-hardened stainless steel spiral band that wipes the column wall with considerable force. This band forces the liquid downward in the column so that a long path of travel with minimum contact time is obtained. With this design, there are no hold-up areas where the material being distilled can remain for any length of time, and, consequently, degradation or poly­merization of the material is kept at a minimum. The

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

128.114.34.22 On: Sat, 22 Nov 2014 15:27:30

NOTES 1003

VARIABLE SPEED MOTOR 50-300rpm

FIG. 1. Falling film molecular still.

strong wiping action of the spiral band is especially effective in removing from the column wall material that has begun to polymerize. The important components of the still are described in some detail in the following sections.

Spiral Band. The band is fabricated of spring-hardened stainless steel ribbon 0.010 in. thick and 0.25 in. wide. by winding by hand on a lathe to give approximately one turn to the inch. The glass tube selected for the still column should be of such a size that the band fits in it tightly. The band is rotated at speeds of 50 to 300 rpm depending on the viscosity and heat sensitivity of the material being distilled.

Pump (A). A simple trouble-free mechanical pump makes it possible to return repeatedly the undistilled residue from the lower reservoir (B) to the upper reservoir (C) without affecting the pressure of the system. The pump comprises a piston of Teflon (tetrafluoroethylene resin) mounted on a polished tungsten rod that operates through a silicone or neoprene stopper. For the final fractionation, the material remaining in tube (D) may be forced up into the upper reservoir (C) by admission of an inert gas through stopcock (E).

U-Ring Seal (F). The udder is attached to the still by an entirely new kind of joint called a U-ring seal. Details of this new type of seal are shown in Fig. 2. The U-ringl is made of neoprene and is held on the smaller diameter glass tubing between the glass rings that are a part of the tubing. This assembly is then inserted into the cup formed on the end of the other piece of glass tubing with the result that a seal is formed on both faces of the U-ring as well as around its periphery. Such a seal is vacuum-tight, very easy to make or break, and eliminates all danger of seizing as encountered in glass joints.

O-Ring Seal (G). The upper part of the still that con­tains the combination seal and suspension bearing (H) for the spiral band is joined to the main part of the still by a modified O-ring seal. This seal employs on both flanges a circular groove of the same diameter as the O-ring. The grooves are semicircular in cross section and about one­third as deep as the O-ring is thick. This arrangement

FIG. 2. U-ring seal.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

128.114.34.22 On: Sat, 22 Nov 2014 15:27:30

1004 NOTES

gives more surface for sealing and always permits perfect alignment in assembling the apparatus.

The combination seal and suspension bearing (H) for the band and the adjustable needle valve (I) for admission of the liquid from the upper reservoir to the still column have been described previously.2

* Contribution':No. 613 from the Central Research Department, Experimental Station.

1 These rings can be obtained from Linear, Inc., State Road and Levick Street, Philadelphia 36, Pennsylvania.

2 R. G. Nester, Anal. Chern. 28, 278 (1956).

Inexpensive High Resolution Wheatstone Bridge

KARL EKLUND*

Wyle-Parameters, Inc., New Hyde Park, New York

(Received May 25, 1960; and in final form, June 27, 1960)

IN order to measure small changes in precision resistors undergoing environmental tests a bridge with very

high resolution was needed, but a standard bridge was not financially justifiable. We therefore constructed the bridge shown in Fig. 1. The variation in balance condition with Rs is

00

X(Rs)/X(O) = 1- (R1/R2) L: [-Rs/(R1+R2)]i. i~l

FIG. 1. Schematic diagram of high resolution Wheatstone bridge.

If R2 is sufficiently larger than Rs this series can be terminated after one or two terms, and an approximately linear variation of balance with Rs is obtained.

In the actual application each of the resistors except Rs was made from a parallel combination of several 1 % deposited carbon resistors, and Rs was a 0.1% linearity multi-tum variable resistor. They had values that per­mitted spanning a 2% range around 129 ohms with a resolution of 0.002%; and a day to day stability of the same order when used to measure the standard resistors for calibration.

The bridge was constructed for less than $50.00.

* Present address: 66 Cliff Avenue, Hempstead, New York.

Vacuum Tube Electrometers Using Operational Amplifiers

G. F. VANDERSCHMIDT

Lion Research Corporation, Cambridge 39, Massachusetts

(Received May 23, 1960; and in final form, July 6,1960)

VACUUM tube electrometers are used to measure current from transducers of the current-source type,

for instance photocells and ion chambers. They can meas­ure a current of 10-14 amp or less. Although excellent commercial vacuum tube electrometers exist, scientific instruments often require built-in electrometers with special characteristics. This note describes two general purpose electrometers using operational amplifiers, that is, packaged differential dc amplifiers with voltage gain of 1000 or more. The use of operational amplifiers results in a saving in design, construction, and maintenance time1

over the adoption of previously published vacuum tube electrometer circuits.2

In the illustrated circuits, the input current develops a voltage across R of JR. The output of the amplifier which follows is 100% fed back to the input of the circuit, and the voltage developed across the meter circuit at the amplifier output is JR. A large enough value of R is used so that J R is large compared to the drift in the input

+16V I IN

10M

R

10K 10K -16V 1.3V

ZERO~ __ -f­SET 10K

10K

+16V

CK 5886 (2)

FIG. 1. Electrometer using Burr-Brown model 1304 transistor op­erational amplifier (Burr-Brown Research Corporation, Tucson, Arizona). RM is adjusted to give full-scale deflection for a maximum value of current to be measured.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

128.114.34.22 On: Sat, 22 Nov 2014 15:27:30