PRANCISJ
34dk [CONTRIBUTION FROM
THE
. NORTON
Voi. 71
RESEARCH LABORATORY OF THE GENERAL ELECTRIC COMPANY 1
The Mass Spectrum of Monoisotopic Bio& BY FRANCIS J. NORTON The dissociation of complex molecules by electron impact presents many features of theoretical and practical interest. This is Particularly SO in the cases where isotopes occur in the molecular structure. The availability of stable separated isotopes from the Atomic E~~~~ commission renders possible the construction Of molecules with altered isotope ratios. This paper is to report the investigation of one of these, B ~ O H ~ , by the spectrometer, and the determination of its molecular fragmentation pattern under electron impact. Normal BzH6 with its pattern was also studied. Normal boron consists of two stable isotopes, about 20% BO ' and 80% B". From the Isotopes Division, U. S. Atomic Energy Commission, was obtained the complexB F ~ . c ~ F ~this . the boron is 96% of the BO ' isotope and 4% of B". Through the kindness of Dr. Edith Boldebuck and Dr. T. R. Elliott of our laboratory, Bi0H6was syntheiized from this enriched isotopic material and was used in these experiments. The mass spectrum of normal polyisotopic diborane has been studied by J. A. Hipple' and by Vernon H. Dibeler and Fred L. Mohler.2 From its mass spectrum they made calculations for the ion distribution of the monoisotopic material. The present work renders possible a comparison between such a calculation and an experimental determination on essentially monoisotopic diborane. Further interest lies in the fact that the mass spectrometers used in references, lv2 and in the present work were made by three different manufacturers. Experimental (a) The synthesis of BioHBwas from the complex BIOFJOC~F~ (the boron consisting of 90%. B'O). This was heated to 350' in a stream of purified nitrogen. The B10F3 evolved was collected in cooled ether and reacted with lithium hydride. The diborane was separated from the ether and subjected to several low temperaturd: vacuum distillations. It was kept stored a t -20 From mass spectrometer analyses higher boron hydrides and normal BsHs, if present a t all, were under 0.05%. Diethyl ether was under O.l%, and a small amount, under O . l % , of SiH,, was also found. Normal B2He was furnished by Dr. L. V. McCarty of our laboratory and was of very high purity. (b) The mass spectrometer used was the 60" type, single focussing with 6-inch radius of curvature. It is manufactured by the General Electric Company. The accelerating voltage was 2000, ionizing voltage 70. The filament current was 4.5 amperes, total emission current 1.0 ma., trap current (electrons collected a t end of ionizing chamber) 0.28 ma. Positive ion currents a t the collector were measured by a 954 preamplifier tube in a direct current feedback amplifier circuit. The initial
.
(1) J . A . Hipple, Jr., P h y s . Rev., 67, 380 (1940). (2) V H nihrler and F 1. Mohler T r r r c l i l1948)
> i i ~ h 4 1
70. Q R 7
voltage drop to ground was through an IRC resistor of 2 x 10' ohms. The amplifier output was recorded on a General Electric photoelectric recorder with appropriate resistances for selecting sensitivity ranges. The range of masses was covered by sweeping the magnetic field. This could be done either automatically or manually. For precise measurement, the desired peak was selected and held a t maximum value on the recorder by manual control. I t was focussed to maximum value by manipulating the focussing, drawing out and beam centering voltages. I t should be pointed out that with this type of instrument, to obtain the maximum peak height as is done with these manual measurements reported , refocussing is needed particularly a t the lower end of the mass scale, for each mass peak. This means that the scan does not represent the maximum height of peak attainable. The peak heights for a given focus do remain reproducible, and the scans in the lower mass regions especially, should be regarded as recognition patterns. On automatic sweep, changes in focus were made for peaks in the 1, 2, 10-13 and 20-28 mass regions, to give them maximum values possible with fevr-est focus changes. The height could be read to * 1% for deflections of the recorder Den from one-half to full scale. For rneasurements, readings were made in the region of 80 to 100% full scale deflection. This could be chosen by using appropriate, known values of resistance in the measuring circuit. The gas to be measured was fed into the spectrometer through an orifice type leak. This was so small that 40 mm. gas pressure on the high side gave 2 X 10-7 mm. on the ion gage a t the spectrometer tube outlet. The gas used during analysis was minute in amount. With that leak, the change in pressure in a 1 liter reservoir was under 2% per twenty-four hours. Meoasurements were made keeping the B&h on the leak a t 25 . I had found in previous work that with very pure, normal diborane handled with good vacuum practice so that the amount of water present is exceedingly small, the following rates of decomposition were observed a t an initial pressure of 33 mm. This extrapolated to 25' op a log OC.
Rate of pressure increase d#/dl (mtn. /min.)
120 110 100
0.045 .020 .005
rate v s . 1 / T plot gives about 10-8 mm./min., or under 0.00570 per twenty-four hours. Before introduction, the hydride was held in a trap at -190" and any traces of gaseous hydrogen pumped away; it was tohen taken over to the leak reservoir with the trap a t -78 . After the mass spectrum was run, the remaining BzHswas taken back into the trap, put a t -190' and any hydrogen measured. I t was under 0.1%. Hence, it is believed the hydrogen ion peaks found are from BsHs dissociation in the miss spectrometer only.
Results The isotupic constitution of the gas molecules entering the spectrometer can be calculated very simply from statistical considerations alone. The general expression from simple probability for the abundance distribution of isotopic species of moleb)"(c d)" where a and b cules is given by (a are the relative abundances of the isotopes of one atom, in this case boron isotopes 10 and 11, and c and d are the relative abundances of the stable
+
+
3485
oct., 1Y49 TABLE 1 ISOTOPIC CONSTITUTION OF THE MOLECULAR GASESUSED "Monoisotopic" BzH! (96% Blo) (4% B") Proportion calculation Per cent.
Normal BZH6 (20% BIOI (80% BL') Proportion calculation Per cent.
Molecular type
0 . 8 X 0 . 8 = 0.64 2 ( 0 . 2 X 0.8) = 0 . 3 2 0 . 2 X 0 . 2 = 0.04
B"B"H6 B1'B"H~ B l*BloHa
0.04 X 0 . 0 4 = 0.0016 2(0.04 X 0.96) = 0.0768 0.96 X 0.96 = 0.9216
64.0 32.0 4.0
0.16 7.68 92.16
hydrogen isotopes 1 and 2. The exponents n and m represent the number of atoms in the molecule. Taking the relative isotopic abundances as : B lo = 0.2, B" = 0.8, H' = 0.998, H2 (deuterium) = 0.02, the expression becomes for BzHB: (0.2 0.8)2(0.9998 0.02)6. For normal deuterium abundances, the second term may be taketi as unity. If for the normal B&& a ratio of B", B'" = 4.0 is assumed, the following Table I gives the constitution for the two types of BZHe. In the mass spectrometer these molecules are broken up by the impact of 70 volt electrons, and form positive ion fragments of various masses. The relative abundance of each of these is given by the positive ion current a t each mass. This current is recorded as peak height, and the mass spectrum is a record of the peak heights or abundance plotted against mass. No way is known to calculate the theoretical abundance of the ion fragmentsfrom complicated molecules. Figure 1 gives a record of the spectra of the two kinds of BzHa used. Masses are recorded a t the bottom, increasing from right to left, and the peak height is the magnitude of the positive ion current. The measured values of the mass spectra are Fig. 1.-The mass spectra of monoisotopic BPHe and of normal B2Hs. given in the following Table 11. I n each case, the dominant peak is taken = 100. and 5 hours. Spectrum d was taken on another day and with another charge of gas. TABLE I1
+
+
MEASURED MASSSPECTRA OF Mass m/c
28 27
Monoisotopic B210Ha Observed in 9SY0 Corrected to BlO material 100% B10 1
.
(2.4)
.
0.5
Xormal BzHs 20% B'Q 80% B" average
(0.1)
-_100.0
26 25
9.4 100.0 -
1.5 100.0 -
96.2 52.5
24 23 22 21 20 14 13 12 11
46.9 28.6 61.0 5.1 1.9
46.5 24.5 63.0 5.1 1.9
10
11.5 45.7 1.6
78.7 38.3 8.5 1.3 0.1 0.5 12.6 8.0 16.0 3.65 71.6 2.4
2
1
...
0.9 9.7 4.6
,..
0 ,5 9.9 4.7 12.0 45.7 1.6
A comparison of reproducibility may be of inwere terest. The following spectra of normal taken; a, b, c on one day after the gas had run 1, 2
TABLEI11 MEASURED MASSSPECTRA OF NORMAL B1H6 Mass
a
b
C
d
Av.
28 27
2.0 100 -
2.0 100
2.0 100 -
3.7 100 -
(2.4) 100
26 25 24 23 22 21 20
95.8 52.3 78.6 37.9 8.6 1.3 0.1
96.5 52.7 79.6 38.7 8.6 1.4 0.1
96.3 53.0 79.5 38.8 8.7 1.3 0.1
96.0 51.9 77.1 37.6 8.2 1.2 0.1
96.2 52.5 78.7 38.3 8.5 1.3 0.1
14 13 12 11 10
...
...
...
12.5 7.9 15.8 3.6
12.9 8.1 16.2 3.7
13.5 8.4 16.3 3.7
0.5 11.6 7.6 15.7 3.6
0.5 12.6 s.0 16.0 3.65
-
-
The measured spectrum for Bi0H6needs correction for the presence of 4y0 B" with the 96y0BO ' in the starting material. How this is done is illustrated by Table IV. Ions having the same constitution as regards the numbers of hydrogen attached to boron are connected by the slanting ar-
FRANCIS J. NORTON
3490
Vol. 71
rows. These ions have similar, presumably identical, probabilities of formation by electron impact.
varied from 4.27' to 4.42depending on the source of the boron. It is conceivable that either the preparation and repeated vacuum distillation of the B2Ha TABLE IV resulted in some enrichment of the light BIO, or CONSTITUTION OF MASSSPECTRUM OF that the probabilities of ion formation involving Mass Constituent positive ions Bl0 and B" are not identical as I assumed. 28 B"B"H6 However, a calculation of sensitivities (that is, 27 BllB1lHj B10BllH6 peak height as a function of the abundance of 26 BllBllH4 BIOBllHs B10B10H6 each type of molecule in terms of its partial presB10BllH4 BIOBIOHa 25 BllBllH, sure on the leak), checks the assumption that the BIOB1lHs BlOBlOH, 24 B1lB1lHs ionization probability is very nearly identical for 23 BllBllH BIOBllHI B10B10H3 ions of similar type as B''B1lH6, B'OB"H6 and 22 BllB11 BlOBllH BIOBIOHz B'OB'OH5, in both the normal and in the monoiso. . . . . . BlOBll BlOBlOH 21 topic material. This is also true for each other ion ...... BlOBlO 20 ,..... type, irrespective of its isotopic constitution. There is one other possibility which could ac14 B"Ha count for the low value of the B1'/Bl0 ratio which Bl0H3 13 B"H2 had to be assumed, compared to the more precise B10H2 12 B"H values of InghramSand Thode, et d4 It may be B1"H 11 B" BlO that the dissociation probability of Bi0H6is greater ...... 10 than that of the heavier molecules. This would 2 H2 give more Bl0 in the dissociation products than in 1 Hi the original gas, and might explain the low ratio Hence, as we proceed downward along any slant- assumed. In calculating the monoisotopic spectrum from ing arrow, say from Bi1H5 mass 27, to B1OBl1H5 mass 26, to Bi0H5mass 25, the theoretical distri- the observed normal spectrum, i t seemed best to bution of these three similar ions can be calculated derive the value for Bll& ion a t mass 28, from the directly, if the isotopic abundances of Bl0 and B" Bi0H6 value, in view of the possible contaminaare known, in the manner shown in Table I. The tions which can accumulate a t mass 28. Hence, probability of Bl1B1l ion formation in the nearly the value of 1.5 was used for the normal B2H6 monoisotopic material is only 0.04 X 0.04 = mass 28 peak height relative to mass 27 = 100. Using the ratio Bl1/B1O = 4.0 for the normal 0.0016. This is too low to be of any importance here, and the whole column of ions from Bi'H, ( x BzHe which I used, a monoisotopic spectrum can be calculated. I n Table V are given: (a) the obfrom 0 to 6) will be taken as zero for this case. The ion a t mass 20 of peak height 1.9 can be served monoisotopic spectrum from B:OHer cordue only to Bio. From its 0.9216 probability, rected for the 4% Bl1 present; (b) the monoisoprojected back up the arrow we arrive a t the simi- topic spectrum calculated from my measured lar ion BloB1l of 0.0768 probability a t mass 21. normal BzHe spectrum; (c) the monoisotopic Its height can then be calculated as (0.0768/ spectrum calculated by Hipple' from his measured 0.9216) X 1.9 = 0.16. Then the measuredmass21 normal spectrum and (d) the monoisotopic specpeak height minus this 0.16 gives 4.94 as the cor- trum calculated by Dibeler and Mohler2 from rected peak height due to the ion BPH only. Pro- their measured normal spectrum. ceeding up the table of constitution, Table IV, in TABLE V this way, we arrive a t the corrected values for MOXOISOTOPIC hfASS SPECrRUM OF RzHs 100% monoisotopic material, Table 11, and put it (d) (a) Obser vtld Calculated (r) on the new basis of Peak 25 = 100 = Bi0H6ion, Dibeler and spr~ctrunl Cdlci1l;lti.d Calculated Ion t h i s papcr t h i q paper Ilipple' M o hler the most abundant ion. In a similar manner, the normal B2& spectrum 1.5 1.5 0 0.64 100.0 100.00 can be calculated to the monoisotopic spectrum, BzHb 100.0 100.0 44.8 52.6 as did Hipple' and Dibeler and Mohler.2 In BzH4 47.0 46.5 26.2 21.7 agreement with the latter, I found that the best BzHj 23.2 24.5 83.7 76.8 agreement with the observed spectrum, and with B:Hr 64.8 63.0 2.9 7.00 4.7 the monoisotopic spectrum here reported, was ob- BzH 5.1 3.43 2.2 9.2 tained by taking the normal distribution of boron Bz 1.9 isotopes as B1l/Bl0 = 4.0. ... 0.62 0.5 BH3 0.6 This is not in agreement with the ratio of 4.311 BHz 19.5 ... 9.0 9.9 determined by Inghram3using BF3 and B(OCH&; BH 10.6 0 5.8 4.7 also T h ~ d e , ~al., e t usingBF3found theB1'/BIOratio B 21.0 40 12.0 14.6
'' '' ''' '' ''
''' 11 ''
(11)
2
'
(3) M. G. Inghram, P h y s . R e v , 70, 653 (1946). F. P. Lossiny and C. U . Collins, THISJ O U R N A L , 70, 3008 (1948). (4) H. G. Thode, J. Macnamara,
H2
H1
45.7 1.6
(45.7) (1.6)
...
2.71 3.98
Oct., 1948
KINETICSOF ACID-CATALYZED HPDROLYSIS OF AMINE DISULFONATE
The concordance, with the exception of Hz ‘ion, is rather good, considering the determinations were made on three different mass spectrometers, manufactured by different companies. The wide differences in hydrogen mass 2 peak heights are undoubtedly due mainly to thermal decomposition in the source region, where there existed different structures, temperatures and differential pumping rates out of the source region, in the case of the three different types of mass spectrometer. The use of the monoisotopic material enables the origin of the small peak at mass 11.5 to be identified unambiguously. This peak was noted by both Hipplel and Dibeler and Mohler.2 The constitution of peak 23 is presented in the following Table VI. The figures are relative to the dominant peak = 100 in each spectrum. I t will be seen that peak 11.5 is proportional to the B1lB1OH2 ion only. Hence, doubly ionized B1lB1oHzion of mass 23 is responsible for this peak. The appearance potential of peak 11.5 was measured as approximately 23 volts (argon calibration), whereas the appearance potential for the
ICOX~RIRUTION FROM
THE
3481
TABLE VI CONSTITUTION OF PEAK23 A-ormal
BzHs
Ion
96% B~H.
Bn”H B”B 10H. B :OH
4.8 31.4 1.4
0 4.7 23.9
Peak 23 height Peak 11.5 height Ratio 11.5 peak to Bl1Bl0H2ion contribution
37.6 0.66
28.6 0.075
}
0.021
0.017
singly ionized peaks was about 15 volts. Very sinall peaks of double ionization also were found a t masses 12 5 and 13.5. Summary The mass spectrum of the monoisotopic molecule @HE was determined. This was compared with the calculated monoisotopic spectrum from measurements on BzH6 of normal isotopic boron ratio. SCHENECTADY, N. Y.
RECEVED APRIL4, 1949
GATESAWD CRELLINLABORATORIES OF CHEMISTRY, CALIFORNIA INSTITUTE OF TECHNOLOGY, NO. 12501
The Kinetics of the Acid-Catalyzed Hydrolysis of Amine Disulfonate Ion; The Third Ionization Constant of Amine Disulfonic Acid’ BY GEORGE J. D O Y L EAND ~ NORMAN DAVIDSON Amine disulfonate ion (imidodisulfonate ion) is known to hydrolyze irreversibly and quantitatively in dilute acid a t a measurable rate according to the equati0n~3~J H+ Hh’(S03)g- f H20 +HSO4- f H~NSOI- (1)
This data is not, however, extensive enough for a kinetic analysis. Because of the lack of quantitative data, we have investigated more thoroughly the kinetics of reaction (1). Materials.-The distilled water, standard base (0.02 N sodium hydroxide) and standard acid (0.02 N hydrochloric acid) were prepared and stored so as t o be free of carbonate or carbon dioxide. A solution of sodium chloride used t o adjust ionic strength was prepared from the C. P. salt which was dried four hours a t 2 0 0 O . 7 A solution of sodium sulfate used as a source of sulfate ion was prepared from the C. P. anhydrous salt which was ignited one and one-half hours a t 700-800”.* Potassium amine disulfonate was prepared by hydrolysis of potassium amine trisulfonate
Wagner6 in the course of his study of the kinetics of the hydrolytic decomposition of the sulfur-nitrogen acids did one experiment on the hydrolysis of amine disulfonate in dilute hydrochloric acid solution. On the basis of this meager evidence he concluded that the rate of hydrolysis is proportional to the product of the concentrations of hydrogen ion and amine disulfonate ion, and is H+ N(S03)s’ f H2O --f HN(S03)2’ + HSOI- (2) therefore autocatalytic. Sisler and Audrieth4 have presented valuable data on the hydrolysis of This in turn mas prepared by reaction of a large excess of amine disulfonate and on the two-step hydrolysis potassium bisulfite with potassium nitrite in hot aqueous ~ of amine trisulfonate ion (nitrilo-sulfonate ion) to solution (Sisler and Audrieth’s m o d i f i c a t i ~ nof~ ~Claus amine disulfonate and thence to amine monosul- and Koch’slO method) fonate ion (sulfamate ion) a t several temperatures. NOz- f 3HS03- --+ N(s03)~-f H2O f OH- (3) (1) A more detailed account of this research is contained in t h e thesis by George J. Doyle, submitted in partial fulfillment of t h e requirements for the degree of Master of Science a t t h e California Institute of Technology. (2) Present address: Department of Chemistry, University of Indiana, Bloomington, Indiana. ( 3 ) Raschig, A n n . , 241, 161 (1887). (4) Sisler a n d Audrieth, THISJOURNAL, 60, 1947 (1938). ( 5 ) Audrieth, Sveda, Sisler and Butler, C h o n . Rev., 26, 49 (1940). ( 6 ) Wagner, Z . phrsik. Chrm.. 19, 678 (1896).
A preparation on a scale of about ten times greater than that used by Sisler and Audrieth gave a yield of potassium amine disulfonate, washed free of sulfate, of 64% based on potassium nitrite; Sisler and Audrieth report 51y0. The (7) Sorensen, Z. anal. Chem., 44, 140 (1905). (8) Remy and Siegmund, ibid., 93, 321 (1033). (9) “Inorganic Syntheses,” McGraw-Hill Book Co., New York N. Y., 1946, Vol. 11, p. 182. (IO) Claus and Koch, A?,%.,162, 336 (1860).
