Table 1. Cell Length Ratio Measured over Useful Concentration Range of Cr as Crz07-2
bulbs 0 9965 0 9995 0 9991
1 0007 9986 9990 9987 9968 9983 9985 9971 9991 9967 9968 9975 9975 9998 9989 9990 9982 0.9979 Av. 0 9983 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
[Cr] in sample cell in mg. /liter 137 66 123 28 107 04 91 245 74 267 57 266 57 001 53 492 48 351 43 159 38 932 37 596 31 712 25 952 20 055 16 369 lt5 305 13 850 13 500 12 353 10.787
scribed with this cell in the light beam. Stock solution was then added to the other cell from a 5-ml. microburet, and yo T readings were taken when the concentration of the standard cell approached and passed that of the sample cell. A graph was drawn (see Figure 1) and the concentration in the standard cell was determined when the readouts ,!'ere the same for both ce113. Under the conditions employed the % T reading for the sample (set point) tended to drift. The value of the set point had to be checked often and slight adjustments were made to keep it constant. RESULTS AND CONCLUSIONS
The ratio of bulb, was calculated from the data obtained as described over a concentration range of 137.7 to 10.79 mg. Cr per liter (see Table I). This ratio was 0.998 with a standard deviation of +1.13 p.p.t. and a 95% confidence interval of p.p.t. Using the two cells and the method described it is possible to determine dichromate in the
concentration range to 10-4Jl to an accuracy of better than 2 p.p.t. For substances with molar absorptivities greater than lo4, this method would allow determinations to an accuracy of 1 2 p.p.t. tc; concentrations of 10-6.U with an unmodified Beckman Model B spectrophotometer, provided a suitable standard solution was available. The method provides maximum scale expansion for every sample used, requires only one stock solut,ion, and provides a comparison measurement for increased precison. The only disadvantage encountered is the need for the rather bulky equipment and time necessary for the isomation. Precision and ease of handling could be improved by the use of glass cells instead of the easily scratched and difficult to clean plexiglas cells. LITERATURE CITED
(1) Lingane, J. J., Collat, J. W., CHEM.22, 166 (1950).
.~N.IL.
(2) Reilley, C. K., Crawford, C. M., Zbzd., 27, 716 (1955).
Magnetic Tape Recording of Analytical Data Phillip Issenberg,' M. L. Bazinet, and Charles Merritt, Jr., Pioneering Research Division, U. S. Army Natick Laboratories, Natick, Mass.
erritt (2) has reported the use of magnetic tape for recording mass spectra derived from a rapid scanning time-of-flight mass spectrometer, and concurrently with this communication the recording of fast scan high resolution mass spectra is reported ( 3 ) . This communication describes the technique employed for recording multiple fast scan mass spectra of compounds separately eluted from a gas chromatographic column and elaborates some of the advantages of magnetic tape in modern analytical data acquisition and processing. The versatility of magnetic tape is a consequence of the fact that electrical signals are recorded essentially in electrical form. They may be reproduced a t some future time a t faster, slower, or the same tape speed used for recording the signal. This time scale expansion and compression feature can facilitate interfacing of analytical instruments to data acquisition systems which provide digital records suitable for computer processing. Data recorded in a variety of time frames may be matched, via analog magnetic tape, to a single digitizing system, reducing the complexity and cost of such systems. XIagnetic tape recorders are available with a variety of features, degrees of sophistication, and costs. The specific 1 Present address, Department of Ktltrition and Food Science, llassachiisetts Instittlte of Technology, Cambridge, Mass.
1074
ANALYTICAL CHEMISTRY
instrument needed obviously depends on the requirements of the laboratory. Most instrumentation class recorders offer tape speeds of 1 7 / 8 , 33/4,7 l / 2 , 15, 30, and 60 inches per second (i.p.s.). Tape speed accuracy between 0.2 and 0.5%, and flutter (short term tape speed deviation) between 0.2 and 1.0% are typical of lower price instrumentation recorders. Since recorders may be purchased with seven to fourteen available data channels, it is usually feasible to utilize one channel for recording an accurate timing signal. Oscillators or pulse generators of any frequency and degree of accuracy may be employed depending upon the timing accuracy required in analysis of the data. Thus, an internal standard is provided which makes time measurements essentially independent of tape speed. Independent variables other than time, such as magnetic field strength, voltage, or temperature may also be employed in this manner. Full scale signal to noise ratios of 40 db. (1OO:l) are typical of instrumentation recorders operating in the F M mode. This limitation on the dynamic range of the recording system is not serious if the multiple track recording capability is utilized with each track rereiving a signal a t different attenuation or gain level in a manner similar to that employed with mass spectrometer oscillographic recorders. High speed automatic attenuators may be
used to extend the dynamic range for single channel recording. Magnetic tape recording alleviates many of the data handling problems associated with mass spectrometric analysis of gas chromatographic effluents. If one wishes to identify the maximum number of components of a complex mixture of volatile compounds, a large number of spectra-e.g. 200 to 300-must be recorded during the chromatographic separation. Often many preliminary preparation steps are required to obtain a suitable sample; therefore, it is desirable to extract as much information as possible from a single sample. The probability of missing a significant compound should be minimized. These Considerations require either careful judgment by the operator as to when to record spectra during elution of chromatographic peaks, or continuous scanning of spectra during the entire course of elution. If magnetic tape is used as a recording medium, repetitive mass spectral scanning may be employed and all data are permanently recorded in convenient form. The tape may be played back and spectra may be examined carefully prior to permanent recording in any desired format. n a t a may be examined using an oscilloscope triggered from a scan start pulse recorded on one tape channel, or a recording oscillograph at a slow chart speed may be used.
MASS Figure 1.
NUMBER
(m/e)
Mass spectrum of hexadecane
Upper trace: Direct oscillographic record Lower trace: Reproduction o f osiilographic record of spectrum recorded on magnetic t a p e Galvanometer frequency response, 1 kc. Oscillograph chart speed, 8 i.p.s. Spectrum scon rote, m/e 12-200 In 6 sec. Tape speed, 7’/2 i.p.s.
Figure 1 illustrates the accuracy with nhich mas5 spectia may be reproduced by magnetic tape recording. The electrometer output of a Bendix Timeof-Flight mas1 spectrometer was recorded by both an oicillograph and a magnetic tape recorder (CEC Model VR 3300) operated in parallel The tape recorder output aas displayed on a second oscillograph trace. Displacement b e h e e n the t n o hexadecane spectra represents the time required for the tape to move from the record head to the reproduce head. Only one channel of data was recorded in the interest of clatity The tape \vas recorded and 1.p.s. reproduced in the F M mode a t Owllograph chart speed \\at 8 1.p.s. When differences appear, chart speed may be increased to examine details of the spectrum. Figure 2 Mas taken from the same tape recorded hexadecane
L
1
MASS
NUMBER
[m/e)
Figure 2. Oscillograph record of five repetitive scans of a hexadecane spectrum recorded and reproduced on magnetic tape at 7l/2i.p.s. Galvanometer frequency of response, 1 kc. Oscillograph chort speed, 0.1 25 1.p.s. Spectrum scon rate, m/e 12-200 In 6 sec. Timing signal, 250 p.p.s.
I MASS Figure 3. Portion of hexadecane spectrum ( m / e oscillograph
NUMBER
57-99) recorded
(m/e)
on magnetic tape at
7l/2 i.p.s.
and reproduced on an
Galvanometer frequency response, 1 kc. Oscillograph chart speed, 8 1.p.s. Timing signal, 250 p.p.1.
VOL. 37, NO. 8, JULY 1965
1075
spectra as Figure 1, but the oscillograph chart speed was 0.125 i p s . At this speed a large number of adjacent spectra may be examined for detection of differences. Figure 3 illustrates the use of this technique for closer inspection of the m l e 57 to 99 region of the same spectra at a chart speed of 8 i.p.s. The bean rate for all spectra was 6 see. for a range of masses from m / e 12 to 200. The reference timing signal seen a t the bottom of each record was generated by a General Radio Type 1217-B pulse generator with a repetition rate of 250 p.1, .s.
A chromatogram may be displayed either by recording total ionization on one tape track or by playing back the
tape into a slow response servo-recorder in parallel with the oscillograph as described by Dorsey, Hunt, and O'Neal (1). The latter method provides greater economy of tape track utilization without loss of data. Magnetic tape provides a simple versatile means for recording large volumes of analog data which must be examined in detail a t some time after recording. When the data have been transferred to graphic form or digitized, the tape may be re-used or stored for permanent reference. -4more detailed description of a completely automated data handling system for fast scanning mass spectrometers will be the subject of a forthcoming manuscript.
ACKNOWLEDGMENT
The authors gratefully acknowledge the assistance and advice of Robert Olweiler and Edward Mortenson of Consolidated Electrodynamics Corp.
LITERATURE CITED
(1) Dorsey, J. A., Hunt, R. H., O'iYeal, nf. J., ANAL.CHEM. 35,511 (1963). ( 2 ) hlerritt, C., Jr., Paper presented before 3rd Annual bleeting of ASTlI
Committee E-19, Houston, Texas, October 1964. ( 3 ) Merritt, C., Jr., Issenberg, P., Bazinet, M. L., Green, B. X.,Merron, T. O., Murray, J. G., ANAL.CHEM.37, 1037 (1965).
X-Ray Crystallinity of Polymer Emulsions Bert H. Clampitt and Charles
E. Walker,
vc
Thile t,he emulsion polymerization of polyethylene has been known for some time, the characterization of t,he emulsion per se has received only limited study. The analysis of the emulsion is usually limited to the determination of particle size by light scattering or by soap titration. Properties of the polyethylene such as molecular weight, density, crystallinity, glass transit,ions, etc., are usually determined on the polyethylene after it has been isolated from the dispersing medium and these properties may or may not be t'he same as the polymer dispersed in the emulsion. Of particular interest in the present study is the assumption that the crystallinity of the polymer in the emulsion is the same as that of bulk polyethylene. This assumption is usually made in the calculation of particle size from light scattering data, where the refract'ive index of the polyethylene particles is taken as 1.52. If, however, the emulsion polymerization process were to produce an amorphous polyethylene emulsion, then the value to use for the refractive index of the dispersed particles would be 1.49. If the Rayleigh equation (6) is used to calculate particle size from the light scattering data, then assuming an amorphous polymer leads to particle volumes 18% larger than if one assumes the polyethylene to be partially crystalline. Then, too, one is often curious to know if a different crystal form exists in the latex particles, especially if the material under study happens to be polymorphic. EXPERIMENTAL
Materials. Two polyethylene emulsions, a polystyrene emulsion, and a molding grade polyethylene were investigated. Specifically an anionic 1076
ANALYTICAL CHEMISTRY
Spencer Chemical Division,
Gulf Oil
Corporation, Merriam, Kan.
(Poly-Em 10) and a nonionic (PolyE m 40) polyethylene emulsion were studied. These, as well as t h e molding grade polyethylene (PE-1009), are manufactured by the Spencer Chemical Division of Gulf Oil Corp. The polystyrene emulsion (RWL-112) is manufactured by the Morton Chemical Co. -411 of the emulsions were 40YG total solids and they possessed particle sizes in the range of 500-1000 angstroms. S o t only were the emulsions per se studied, but the two polyethylene latexes were further characterized by determining the x-ray diffraction patterns on solids derived from the emulsions. Two methods were used to prepare solids from the emulsion. The first method was to freeze-dry the emulsion which results in a fine powder which could be analyzed directly by conventional x-ray techniques. The second method consisted of coagulating the latex with acetone, filtering, drying, and making a compression-molded specimen for x-ray analysis. Procedure. A standard aluminum specimen mount for the Norelco diffractometer is prepared by painting a thin layer of P E R M A - T I T E (Permatite hffg. Co., Minneapolis 13, hlinn.) liquid rubber adhesive around t h e opening in the mount and on the upper side. When the solvent has evaporated from this material it becomes a pressure-sensitive adhesive and is used to hold the very thin collodion window firmly in place. The collodion window which is from 10 to 20 microns thick is prepared in the following manner: 2 or 3 ml. of collodion is placed on a flat glass plate using an eye dropper. & i napplicator, made by wrapping two or three turns of cellophane tape at two different places about 4 inches apart on a glass stirring rod, is used to draw the collodion into a thin sheet. b'hen enough of the solvent has evaporated to leave a solid collodion film which is still firmly adhered to the glass plate,
the aluminum specimen mount is pressed, tacky side down, on a clearlooking area of the collodion film. Upon thorough drying, the film will begin to pull away from the glass plate. At this time, the specimen mount is carefully lifted up with the collodion adhering t'o the adhesive side. The excess is trimmed away from the outside of the specimen mount and the film is pressed into good contact with the adhesive on the mount. With the specimen mount held collodion side down, an eye dropper is used to fill the cavity in the mount with the liquid sample. When the cavity is slightly more than full due to surface tension of the liquid, a microscope cover glass that has been trimmed to the proper size is carefully placed on the surface of the liquid. Capillary attraction will pull the cover glass down firmly on the bottom of the specimen mount, holding the liquid without leaking in the cavity. X-Ray Equipment. The equipment used in this investigation was the standard Norelco diffractometer manufactured by Philips Electronics Instruments. A copper target x-ray tube was used, operated a t 35 kv., 20 ma. 4 xenon proportional counter was used on the wide range goniometer, and t,he angular 28 range transversed by the goniometer was from 9" to 36" a t a rate of 0.5" per minute. RESULTS AND DISCUSSION
I n order to investigate t'he effect of the collodion window, a sample of PolyEm 40 coagulated solids was pressed into a solid block in one of the aluminum specimen mounts and analyzed with and without one of the collodion films. The results showed that although about 15y0 reduction in overall intensity resulted, the calculated x-ray crystallinity (IT-,) is essentially unaffected, the numbers being within experimental precision.