Table I. Analytical Results on UraniumLoaded Graphite Samples
Nature of Nominal % U Sample % Ua Found Graphitized 13.0 13.2 7.5 7.5 12.0
6.35 7.0 8.0 8.0 Values listed are estimates.
Baked 0
12.1
6.0 7.0
Std. Dev. 0.20 0.08 0.10
0.06 0.10 0.07
been destroyed, cool the solution. Dilute and filter if necessary to remove silica. Transfer the solution to a 50-ml. volumetric flask and dilute it to the mark with water. Read the absorbance of the solution us. water a t 420 mw. Prepare calibration standards from reagent grade uranyl nitrate by the above method. Standards should cover the range of concentrations 1.0 x 1 0 - ~ to 2.0 X 10-3M as DISCUSSION A N D RESULTS
carbide disintegrates the sample into a suspension of fine particles of graphite in a solution containing most of the uranium. Conditions are then excellent for complete osidntion of the graphite. METHOD
REAGENTS. Perchloric acid, 70t072y0; nitric acid, concentrated; sulfuric acid, concentrated; uranyl nitrate hesahydrate, reagent grade. APPARATUS.Erlenmeyer flasks, 500ml.; spray traps and fume collectors as specified by Smith ( 4 ) ; volumetric flasks, 50-ml.; a colorimeter or spectrophotometer. ANALYTICAL PROCEDURE. Prepare the mixed-acid solution using equal volumes of nitric, perchloric, and sulfuric acids. Weigh the sample, containing 0.05 to 0.40 gram of uranium, into a 500-ml. Erlenmeyer flask. Add 25 ml. of the mixed acids, place a spray trap in the flask, and insert it into a fume collector. Heat the flask with a suitable hot plate until white fumes are evolved. Replenish perchloric acid as necessary with occasional small additions of 5 t o 10 nil. When all of the graphite has
Silverman and hIoudy ( 2 ) have published a direct colorimetric determination of uranium in perchloric acid solution. They specify remora1 of sulfate before preparation of the colorimetric solution; however, Scott and D k o n ( I ) had shown that a considerable increase in sensitivity could be obtained by adding sulfuric or phosphoric acid to aqueous solutions. Scott and Dixon’s data show that colorimetric sensitivity reaches a constant maximum value above the 10% (w./v.) sulfuric acid concentration level. It would, thcrefore, appear advantageous to add sufficient sulfuric acid to thc colorimetric solution to reach this plateau, rather than to attempt to remove sulfate from the sample. As sulfuric acid was found to incrcase the rat? of oxidation of the graphite pieces, the amount used in the present procedure was chosen to correspond to a final concmtration greater than 10% after osidation and dilution. An absorptivity of 1.38 X lo3 is obtained using the mixed sulfuric and perchloric acid system, compared with a coefficient of 1.13 X lo3 calculated from Scott and
Dixon’s data. Because esccss sulfuric acid is present, no loss of accuracy would be expected to result from fluctuations in sulfate content of the samples. It is impossible to obtain what might be called a “primary standard” uraniuniloaded graphite; however, a green mi.; containing graphite, carbon, and bituminous material was carefully prepared. Results of analyses of this mix were as follows: calculated U2% content, 14.00%; Uz%found, 14.04 + 0.02%. Results obtained from analyses of baked and graphitized samples are summarized in Table I. The technique has been used unchanged on samples containing uranium of various isotopic compositions. Experience has shonn that the inhomogeneity of the graphite pieces is the largest contributor t o variation in the prccjsion of the mcthod. ACKNOWLEDGMENT
Samples were preparcd by T. 11. Bcnziger, R . K. Rohwer, nnd K. H. Stcin of this laboratory. LITERATURE CITED
(1) s c o
462
(1941).
RECEIVEDfor review May 4, 1959. Accepted September 28, 1959. Work performed under the auspices of the U. 8 Atomic Energy Commission.
Mass Spectrometric Analysis Rearrangements in Vinyl Derivatives FRED W. McLAFFERTY Easfern Research laboratory, The Dow Chemical Co., Frarningharn, Mass. .The typical rearrangement found in the mass spectra of polar unsaturated compounds can also explain maior anomalous peaks in vinyl ethers, vinyl esters, and olefins. It accounts for almost all base peaks of even m/e in olefin spectra. Rearrangement of one (and sometimes two) protons through a six-membered ring intermediate i s used to interpret the effects of substitution.
C
of a bond beta t o a polar unsaturated group (carbonyl, nitrilr, phosphate, ctc.) accompanied by LEAVAGE
2072
ANALYTICAL CHEMISTRY
the intramolecular rearrangement of a hydrogen atom is a prominent feature of the mass spectra of compounds containing such substituents (10). Several chemical analogies HZ$\f+ H2C-, CH;
C
‘R
Hz; HzC
H+
+
phase unimolecular eliniination reactions of this type, such as 0 R-CH~-CH~-O!OR~
-
R-CH
= CHZ
+
ii
HOCORs
(2)
I C+R
(1)
to such rearrangements are found in thermal, photochemical, and high energy radiation reactions (IO, I 4 ) . Maccoll has recently discussed (8) some gas
He postulates a mechanism analogous to Equation 1, and points out the surprising similarities to chemical rcactions in polar solvents. He also draws attention (8) to the similar vapor phase pyrolysis of vinyl ethers to yield carbonyl compounds and olefins.
n
Table 1.
Prominent Ions from Vinyl Ethers,
(3)
Compound Parent EXPERIMENTAL
Uiilesb otherwise noted, the spectra
2.2
56
(6) were obtained on 90”-sector mass
d+ 1
+
42
(,)
c+
/J+
4.2a
81a
1000
13a
c
loo=
25
4+ 320
spectrometers described previously (IO). Inlet temperatures were 200” C . 23
c
c
c
VINYL ETHERS
This rearrangement mechanism will also account for major anomalous peaks in the mass spectra of such vinyl ethers. Thus, t’he largest peak of ethyl vinyl ether is at d e 44, corresponding in mass to CH3COH by t’he rearrangement shown in Equation 3. The prominent peaks of the available vinyl ether spectra are shown in Table I (6). The common cleavages of saturated cthers, RCH20CI12R,would predict ( 9 ) ions as RCH2+ (columns 4 and 9), and RCH20CI12+ (column 3). Cleavage a t the allylic bond yields the oxygen-containing ion of column 6. There also appears to be some tendency to yield the olefin ion (column 5 ) through this allylic cleavage and loss of hydrogen, similar to the olefin ions formed in esters (16) and alcohols ( 3 ) . The peak corresponding t o the rearranged oxygenated ion of Equation 3 (column 7 ) can for several compounds also arise from other breakdown mechanisms. Some are of small probability, however, such as the cleavage of the vinylic bond and loss of hydrogen necessary as an alternate explanation of the m, e 44 of ethyl vinyl ether. The few examples available indicate that Reaction 3 is less favored over the formation of the nonrearranged allyl ion 0--CH=CH2’ when the accompanying neutral product is stabilized by alkyl substituents-i.e., propyl and highrr vinyl ethers. In the spectra of esters (10, IGj, two hydrogen atoms can be rearranged on cleavage of the beta bond. This is also found in the vinyl ethers (column 7 ) , although to a less striking extent. The probable structure of the doubly rearranged ion from the vinyl ethers, CH3CR=OH+, would not have the resonance stabilization of the corresponding ester ion, HOCR=OHT (10). The acetal ether ( 7 ) of Table I also yiclds major ions a t m ; e 74 (base peak) and ?n e 75 (57% of the base peak). These ions might arise through allylic rearrangement of this typr to an mtrr ion, follon.ed by the usual cstc-r tl(,composition with rearraiigcmcnt of on(’ o r two Iiydrogcn atonic (IO,I O ) .
24
1.8
20
100e
20
1006
79“
28
52s
G9a
69.
93a
11a
35.
11
27”
8
798
796
55a
18“
21
6
410
11
16
57
25
2.3
IO“
11
c
c
c
100
15
31
11
18
0.4
100
100
5.2
17
32
6.7
2.4
29
81
Numbers represent per cent of base peak intensity. Contributions froni not been removed. a Minor part may be due to other degradation mechmiiams. b Purity unknown (7). c Not determined. Reference ( 1 ) . e Major part can be due to other degradation mech:uiisnis. m/e 94 = 16, m / e 96 = 5 . 2 .
Table II.
Vinyl Esters
Vinyl Vinyl IsoproVinyl l’ropion- Butyr- penyl n~,’? Acct:tte ate ate Acetate“ 43 100 13 100 44 4 8 11 9 3 45 0 5 0 1 2 1 8 57 2 1 14 58 0.5 59 Values represent per cent of base peak
intensit!.. Reference (f2).
1 2
10
11
3.5 C13
1s
and HZhave
“beta” cleavage with single hydrogc*ii rearrangement ( 5 , f 0, f 7 ) . As found ivith the analogous rcarrangement involving carbonyl, ctc.. groups ( I O ) , tlic six-nicnibered riiiy internicdiate must be possilde structurally to permit the 11yclrogc~11 migration. Thus, methyl ethers are ttio small and no such rearrangcmcnt ifound. A R O M A T I C ETHERS
5
Similarly, the allylic cleavage ith rearrangement of the x+~yl ether RCH?CH,OCIt’=CHR’’ would give O=CR’ -CH2R”+. If R’ is n-propyl or larger, or if R ” is ethyl or larger, this carbonyl ion could then be capable of undergoing further loss of C,H4 or higher oiefin hy
Plicnetole can Ix viewved as a i h y l ether system
ant1 it givcs the rcarrangcd ion C ‘ J f 6 0 T as its largtist pcak ( I 0). Corwspondiny rearrangcwients are eharactcristic base or major peaks in the spectra of phrn~.l ethers (alt’hough C,H,iL”CGHs g i \ w CeH7X’ i n only 1.4y0of the hasc y a k intensity). The presence of the plit~nyl group offers further mechanistic opportunitics over the simple vinyl ctliclr VOL. 31, NO. 12, DECEMBER 1959
0
2073
case, so that the possibilities of pibonding of the transferred hydrogen to the ring, or ion structures as
as well as the analogous
!XI+ 0
should be considered. Transfer of the hydrogen to the oxygen through a fourmembered ring intermediate to yield the phenol ion directly seems improbable because the phenyl group should lower the electron density on the oxygen atom. This enhanced effect of the aromatic ring is also seen in the prominent corresponding rearrangement ion C,HB+ [possibly the tropylium ion ( I 5 ) ] from both H3COCH2C& (10) and HOCH&H&&, (4, IO), while allyl ethers do not show appreciable amounts of this rearrangement, p Diethoxybenzene (molecular weight 164) gives its base peak at m/e 110 corresponding to two rearrangements, each with the loss of C2H4. I n addition, it yields the expected prominent rearrangement ion from loss of CzHl a t m/e 138. ESTERS
The -O-CH=CHZ group can cause rearrangements in other types of compounds. Newton and Strom (IS) have noted the prominent rearrangement ion of m/e 58 in isopropenyl acetate (Table 11). They postulate its formation through alpha hydrogen transfer to yield the acetone ion in a very similar fashion to Reaction 3 where CH3C H r 0 C R = C H 2 is replaced by CH8CO-OCR=CH2. Table I1 also includes the spectra of three vinyl esters (6), whose rearranged ions are considerably reduced in magnitude compared to the previous vinyl ethers. Here COO+ and COOH+ could contribute to the m/e 44 and 45, respectively. Also, rearrangement is probably less favored than in the vinyl ethers, as the neutral products formed, RCH=C=O (ketene) and RC=C=O, would be less stable than the olefins from the vinyl ethers, and for the double rearrangement there are no hydrogens next to the cleaved bond (11). The enhancement of the rearrangement by the methyl of the propenyl ester is similar to effects found in olefins, and will be discussed later in this paper. Phenyl esters show this rearrangement much more markedly than vinyl esters, as WM noted with the ethers. Thus, phenyl acetate has its base peak at m/e 94, corresponding to the loss of CH2CO. There is no evidence for the rearrangement of the second hydrogen atom (m/e 95:m/e 94 = 0.067). I n
2074
ANALYTICAL CHEMISTRY
both . m-ethylphenyl acetate and p ethylphenyl propionate the two largest p e a b in the spectrum are a t m/e 122 and 107, the former as expected from this rearrangement mechanism. The structure of the m/e 122 ion may be similar to that of the molecular ion from ethylphenol, as m/e 107 is the largest peak in its spectrum. OLEFINS
Applying such rearrangement mechanisms to vinyl compounds without the adjacent oxygen can explain prominent peaks in the spectra of aliphatic hydrocarbons. As an example of an analogous chemical reaction, this mechanism has been postulated for the transannular rearrangement of transcyclononene to 1,8-nonadiene by pyrolysis ( 2 ) . The simplest compound where this mechanism could apply, 1-pentene, shows its most abundant mass spectral peak a t m/e 42.
(4)
The only nonrearrangement inode of formation possible for this ion is cleavage of the strong vinylic bond with loss of a hydrogen, which is ordinarily of low probability. Table I11 lists the available spectra ( I ) fulfilling the structural requirements for this rearrangement showing that this accounts for a number of abundant ions for which no simple cleavage mechanism can be ascribed. The double rearrangement of hydrogen, however, appears not to take place. This hydrogen rearrangement with allylic cleavage on substituted l-pentenes seems to be favored by substitution in the 2-position (as noted above for esters), but is lowered by substituents in any other position. A Zalkyl group should increase the electron density a t the double bond, so that the enhancement of rearrangement indicates that the electron pair transfer is from the olefin to the new C-H bond (Equations 1, 3, and 5), as proposed for analogous chemical reactions (2, 8). Transfer in the opposite direction was formulated previously (10). Alkyl substitution in the 1-position should also increase the electron density a t the double bond, but steric interference with the hydrogen migration evidently overcomes this effect. The similarity of the spectra of the cis- and trans-2hexenes must then indicate isomerization in the activated molecular ion or transition state. Disubstitution in the 1-position almost eliminates the rearrangement, even when a %alkyl group is present. The lowered abundance of rearranged ion with the 4 or 5-alkyl substituents may be partly due to the increased
tendency after allylic cleavage for the positive charge to go to the fragment away from the double bond. The secondary and tertiary carbonium ions thus formed would be more stable than the primary ions from the unsubstituted olefin. Possibly the 3- and 4-branching would stabilize the ion from either fragment sufficiently to decrease the hydrogen rearrangement part of the reaction. Of the. olefin spectra available, 5-methyl-1-hexene (base peak a t m/e 56) is the only one in which an odd-electron (even mass number) ion is the largest in the spectrum and is not accounted for by this rearrangement. This may be due in part to the charge going to the opposite fragment through such a mechanism. In the analogous
ester rearrangement, all the isobutyl esters reported by Sharkey et al. (16) show strong m/e 56 ions. I t is interesting that of the 235 noncyclic hydrocarbon spectra available ( I ) , only two (ethane a t m/e 28 and 3,4dimethylhexane a t m/e 56) besides those explained above give an oddelectron ion as their highest peak. There is some evidence for the postu1aLed rearrangement mechanism in the metastable ions of these spectra, though many of the olefins show no such ions from any degradation path. Thus, in 1-pentene, m/e 25.1, postulated as 29+ + 27+ 2, might also arise from 28 (m/e 25.2). I n 270+ 4 42+ hexene and 2- and 3-methyl-1-pentene, m/e 37.1 may be partially 84+ 56+ 28 (m/e 37.4). In 4octene, m/e 63.0 arises from 112+ + 84+ 28. In RCH2CH2CH2CR'=CHR", if R' is n-propyl or larger, or if R" is ethyl or larger, it is possible that two such rearrangements can take place in series, similar to that postulated for vinyl ethers. Thus, the strong peaks at m/e 98 and 70 in 4-ethyi-3-octene (1) might arise from
+ +
-
+
+
w'-1
+
r t , . i 1 9 8 1
Et
Et
+
+
When the intermediate ion is less stabilized, isomerization may take place before the rearrangement. Thus, slthough Pnonene (1) shows appreciable m/e 84 and 42 ions aswould be predicted, the metastable peak m/e 37.3 indicates the reaction 84+ -t 56+ 28 also occurs. Rearrangement of hydrogen also occurs on beta cleavage of acetylenic compounds. Thus, the second largest peak of 1-pentyne is C a d + (largest is
+
Cs137T). 4-Octyne yields appreciable peaks a t m/e S2 and 54, possibly from two such rearrangements in series. 82+ 28 The bieskdown 110+ is c o n f i r m ~ lby a metastable ion at m 'e 61.1. In 5-;lwyiie, the largest peak in the spectrum, rn 'e 54) could arise from two such coiisecutive rearrangemerits 1 u t tlic magnitude is surprising. I t 19 of intewst that, although aliphatic hylrocarboiis tend to undergo i andom tvpe rcarrangements (IO), the addition of just a double bond to such a molcclilc. can rause a major specific rearrni~jit~iiicwtion. Thii ~ncclianism can also be extended to explain the prominent rearrangement ions in alkylbenzene spectra if,D)-e g., n-decylbenzene has its h e peak a t m 'e 92 /C7H8).
-
Table 111.
+
Beta Cleavage with Rearrangement in Olefins
Compound
't" ACKNOWLEDGMENT
lA
The author is indebted t o H. H. Freetlmm for most helpful discussions on the theoretical aspects of this work, and to R. S. Gohlke and V. J. Caldecourt for making their mass spectrometric fles and facilities freely available.
V
33
45
100"
29
100
755
59"
11
72 100
334 39=
100"
30 32
2.6"
31
100
26*
30
100
23b
17
100
52a
1.7
82
12"
31a
5.0
100
10
4.4 42
42
100
4. I
9.7"
100
66a
18
1000
4.9 9.6a
100
94a
9.8"
2.8
3.2
M
7.8
22
loo=
9.6"
K
9.0
100
120
2.9"
17
78
215
1.10
14
93
29
1.9
20
100
1O b
0.4
25 28
100 56
13
11
93
17
13 100 100
24 6.4
0.2 0.6
6.9 65"
1004
8.4
100
0.7
15b
0.3 0.4
7.8
6.2 3.3
0.2
7.6
83
38"
100"
26
100"
225
14
11
0 n - h
6.9 20 20
9 '
16
n-cku
"-CIu
-,y-Cs
-
4.1a
4.3 1.0 0.9 1Ga
20 29 25 23
(12) lleyerson, Seyiiiour, ~ p Spectrosp ~ C O D U 9., 120 i1955). (13>'y:cn?on, A. S.; Strom, P. O., J . Phys. < h c r ? ~62, . 25 i19.58) (14) S i c h o l m n , ;\.-*i.' C., Trans. Faraday c ;OC. 50, 1067 (1954'. (15) I!ylander, 1'. Y . , lfeyerson, Seymour,
@ + J
54
A . G.. Jr., h h I . CHEX 28, 926 (1956). 14) GiIr)in* J. A . . J . Cham. Phus. 28. 52 (19p8); ( 5 ) Gilpin, J. A., RfcLafTcrty, F. W., ANAL.CHEM.29,900 (1957). (6) Gohlke, R. S., .\lass Spectral Files, Spectroscopy Lab., The Don, Chcniical Co., hlidland, illicli. ( 7 ) I*eRlsnr. It. 13.. bra^. CHEM.30. 1797 ,3398.
5'
1.6
(1) .h Petrol. . Inst. Project 44, "Catalog of Mass Snectnil Data." Carnecie
(1958). (8) .\Iactoll, Alan, J . Chetrz. SOC. 1958,
J+
AJ
LITERATURE CITED
\
Parent
n-C6b,
czlJ~'-c'
10
2.8
1.5 0.3
100
8.9
0.4
100
38"
14
13
lob
1.7
17
13
15
1.3
6.9
84a
Grub\). I h r v . .J. Aim, Chem. SOC.
(10) 'Gii~i:.l.rc,y,
G . , Jr., Friedel, It. A., XS.\I.. C ~ I E S31, I . 87 (1959). (17) Shrkcby, A. G . , Jr., Shukz, J. L., Friedel, It. .I.) IDid., 28, 934 (1956). A\.
Sumbers represent per cent of base peak intensity. Data taken from API file (1). Not corrected for isotopic contributions. a
RECEIVEDfor review July 6, 1959. Accepted October 2, 1959.
Part can be due to nonrearrangement degradation mechanisms.
Part can possibly be due to cleavage of double bond.
VOL. 31, NO. 12, DECEMBER 1959
2075
