Instrument Co.) to decide whether crystallization occurred within the film. Thirty-seven minutes were required for a single scan. The resolution setting was 980 and that for the pen speed was 7. The water content of the agar slice was determined by weighing after drying in air, in the infrared beam or under vacuum, or over phosphorus pentoxide as indicated. RESULTS AND DISCUSSION
Spectra in agar of glycine and alanine in buffer p H 7.2 and of these amino acids as their hydrochlorides appear in Figure 1. Similar spectra for threonine appear in Figure 2. When the water loss from the agar slices was examined, slices weighing 546.8 and 650.48 mg. lost 482.1 and 520.0 mg. of water 107 minutes after slicing-Le., 65 minutes after slicing plus 42 minutes in the infrared beam. This water loss was approximately 83% of the original slice weight. When these same slices remained in the sample beam
for 42 minutes more, the total water loss was 90% of the original sample weight. When slices with a mean weight of 625.9 mg. ( u = 0.097)' were dried over phosphorus pentoxide the water loss was 585.6 mg. ( u = 0.0245). This water loss corresponded to approximately 93% of the original slice weight. The presence of the 1720 cm.-' absorption band for the hydrochloride but the absence of this absorption in buffer solution relate our spectra to those of Gore ( 2 ) in aqueous solution. In his studies deuterium oxide was used to avoid interference at 1640 em.-' Absorption in this region does not interfere with our demonstration of the 1720 ern.-' carbonyl absorption noted throughout for the amino acid hydrochlorides. Indeed, slices can be dried until very little absorption shows (Figure 1) at 3350, 640, or 800 cm.-', the well known regions of the spectrum where water absorbs. Of course, the reference beam contains an agar film but even then, in all cases, 65 minutes after mounting the sample the amount
of water present finds the film opaque to infrared light. Thirty-seven minutes later, however, the characteristic absorption bands are seen; a t this time the film contains approximately 80 pl. of water. Although all of this water might not be free as solvent, adequate water was present for solvent for the amino acid. The difference between threonine hydrochloride in the crystalline and dissolved states indicates that solution under these conditions of sample preparation changes the absorption of the hydrochloride form of a t least this acid just as it changed the absorption of the dipolar form observed by Gore ( 2 ) . LITERATURE CITED
( 1 ) Feinberg, J. G., Rapson, H. D., Tavlor, M. P., Nuture 181, 763 (1958).
(2) Gore, R. C., Barnes, R. B., Peterson, E., ANAL.CHEM.21,382 (1949). (3) Lauffer, M.A,, Schantz, E. J., Bibchemistry 1 , 658 (1962).
SUPPORTED in part by a research grant (AI-04819) and a training grant from the United States Public Health Service.
System for Vaporizing Samples Directly in Ion Source of a Mass Spectrometer Gerard 1. Kearns,' Research Division, W. R. Grace & Co., Clarksville, Md.
to be analyzed mass must be in the vapor state. Normally, the material under study is vaporized in an inlet system and slowly diffuses into the ionization region. Many inlet systems can be heated to obtain sufficient sample pressure (about lop2 torr) from materials with very low pressures at room temperature (1, 2 ) . The insertion and heating of samples directly in the ionization region reduces the sample pressure requirements several orders of magnitude. The temperature to which the sample must be heated to obtain the necessary pressure, therefore is reduced proportionately. This permits the analysis of many thermally unstable materials. Heating the sample directly in the ion source has many other useful applications. I t makes it possible to obtain maw spectra of materials of very low volatility (3,5 ) . The products from deliberate thermal degradation (4) can be analyzed with a minimum number of collisions. Gaseous products evolved when solid materials-e.g., cataly-tq-are heated can also be detected and identified in this manner (6). This technique is especially valuable when used with a rapid scanninge.g., time-of-flight-mass spectrometer. The sample systems described in the literature ( 2 , 5 ) are not easily adaptable bimERIAL
A spectrometrically
1 Present address, Picker Nuclear Division, Picker X-Ray Corp., Khite Plains, N. Y.
1402
ANALYTICAL CHEMISTRY
to many spectrometers. A new system has been developed which can be used not only with all time-of-flight mass spectrometers, but with any instrument with a n accessible ion source-e.g., sector magnetic field instruments. EXPERIMENTAL
Apparatus. The basic requirement for a n y sample system which enters the ion source region is t h a t the sample be inserted and/or withdrawn without appreciably affecting the low pressure (10-6 torr) of the spectrometer. The system which has been developed utilizes: (a) a linear-motion high vacuum feed-through; (b) a high conductance isolation valve; (c) a vacuum lock pumped by an external vacuum
A:
Normal
Swag*lok
I connlslor
system; and (d) a vacuum-tight sample holder (probe). A linear-motion high vacuum feedthrough is usually the most difficult component to obtain because it must maintain a vacuum-tight connection while the sample probe is manipulated. A simple way has been devised to accomplish this with a standard '/,-inch Swagelok (Crawford Fitting Co., Cleveland, Ohio) connector (Fig. 1.4). These connectors are constructed with a restriction which allows '/(-inch tubing to be inserted only about 3/8 inch. The vacuum seal, however, is made by the compression of ferrules and not by the end of the tubing. The opening in the body of the connector is enlarged (Figure 1B) so the tubing can pass through the entire unit without affecting the vacuum seal (Figure IC). Teflon ferrules are used and they not only make the seal, but also provide a surface for sliding the tubing easily through the apparatus. Figure 2 is a schematic drawing of the apparatus as attached to a timeof-flight mass spectrometer. A '/*-inch ball valve is used as the isolation valve. The vacuum lock [the volume between the connector (a) and the ball valve] is evacuated through the valved vacuum line. The sample probe is inserted through the linear-motion feed-through ( a ) . All connections except the Swagelok nut are silver-soldered to reduce the possibility of leaks. The sample system is installed so the sample can be positioned with 2 cm. of the electron beam. The magnet assembly of the spectrometer used to collimate the ionizing electron beam was modified
S a m ~ l e~ o s i i i o n
/
I
Figure 2.
Sectioned view of sample system
the isolation valve and into position in the ionization chamber. The high voltage and filament of the mass spectrometer are normally turned off during sample maneuvering, but the pressure does not rise above 10-3 torr during this operation. T o remove the sample after analysis, the probe is withdrawn until the heater is beyond the isolation valve. The isolation valve is then closed and the vacuum lock opened by slowly opening the rear vacuum seal.
to permit attachment, of the system and still allow for adjustment of the magnets. The sample probe can be conveniently constructed frcim ‘/.&-inch stainless steel tubing. ]Insulated copper leads are used in the body of the probe to carry t,he current ‘io the heater but do not. produce excessive heating of the probe. The heater i8i made from 7.5niil tungsten wire and is silver soldered or fastened to the leads mechanically. The probe is made vacuuni-tight by sealing the end with a high vacuum epoxy (T’arian Associates, Vacuum Products Division] Palo Alto, Calif.). The sample is placed directly in the heater or in a glass crucible made from ‘/,-inch tubing. About 2 volts (ax.) produce a temperature of 500’ C. There is no interfering: field produced in the ionization chamt’er at. these conditions. Procedure. T h e :.ample to be analyzed is placed in the heater or in a glass crucible which is inserted in the heater. T h e probe is inserted into the vacuum lock, the rear seal tightened, and t,he enfire section evacuated by opening the vacuum valve. After several minutes, this valve is closed and the isolation valve opened. The probe is then pushed through
heating of solids have also been studied using the sample system described. The results have been correlated with differential thermal analysis data in a manner similar t’o that described by Langer and Gohlke (6). The ease with which sample probes can be constructed makes this device more versatile than those previously described (2, 5 ) . For example, the effluent’ from a gas chromatograph can be admitted directly to the source by attaching a valve to the end of a length of stainless steel tubing. A suitable valve is a cross-pattern metering valve (Nuclear Products Co., Cleveland] Ohio, Type 4MX). This valve, when closed to t.he spectrometer, does not affect the effluent pressure. When closed, the probe is vacuum tight. This system can be heated to a t least 150’ C. The mass spectrometer equipped with this device has been modified by installing a baffle to reduce the flow of gas from the source region. This baffle also increases the time required to evacuate the spectrometer after it has been a t atmospheric pressure. The sample system can expedite evacuation by closing the rear seal with a blank. The vacuum system used to pump the vacuum lock can then be used to pump the instrument. ACKNOWLEDGMENT
The author thanks James Rraatz, J. W.Graves, and Fredrick Player for their invaluable assistance in the design and construction of the system described.
RESULTS AND DISCUSSION
The sample system described has been used satisfactorily to obtain the mass spectra of a variety of materials. Figure 3 is the spectrum obtained from the deliberate degradation of polypropylene. The intensities of the CnHSnpeaks show maxima a t C g , CII, CI4, and C n , in agreement with previously reported data ( 7 ) on the pyrolysis of polypropylene. The sample was heated to approximately 400’ C. The deliberate degradation of copolymers has been used succe.sfully to identify the specific monomeri (Figure 4). The products evolved during the
LITERATURE CITED
(1) Beynon, J. H., “Mass Spectrometry
and Its Application to Organic Chemistry,” Chap. 5 , Elsevier, Amsterdam, 196?. ( 2 ) Biemann, K., “Mass Spectrometry, Organic Chemical Applications,” Chap. 2, McGraw-Hill, Kew York, 1962. ( 3 ) Ibid., Chap. 10. ( 4 ) Brandt, P., Dibeler, V. H., llnhler, F. I,., J . Res. S a t l . Bur. Stds. 50, 201 (1953). ( 5 ) Gohlke, R . S., Chem. I n d . 1963, Yo. 26, p. 946. (6) I,anger, H. G., Gohlke, R. S., ASAL. CHEM.35, 1301 (1963). ( 7 ) Schooten, J. V., Wigga, P. W . 0 . ) SOC.Chem. I n d . ( L o n d o n ) Monograph No. 13, p. 432 (1961).
CHp~C-COOCH,
I
r
n
4
5
7
8
Carbon
9
0
I1
12
13
14
15
16
17
I8
Number
Figure 3. Spectrum of degradation products from polypropylene
~
0
20
40
m/e
60
IO0
80
Figure 4. Spectrum from deliberate degradation of a copolymer (syrene-methyl methacrylate) VOL. 36, NO. 7, JUNE 1964
1403