1
Supplemental Material
2
Additional mass spectrometric methods
3
Instrument configuration
4
All of the measurements described here were made in dual-inlet mode, meaning
5
approximately equal pressures of sample and standard gases are introduced into the
6
bellows of a dual inlet system, and the capillary bleeds and changeover valve block of the
7
Ultra are used to alternately introduce one and the other into the ion source. Typical
8
bellows pressures are ~10 mbar and typical source pressures are ~6x10-8 mbar (both
9
values are background corrected). These values provide a successful compromise
10
between the desire for high ion-beam intensity (so acceptable counting statistics can be
11
reached in a reasonable time) and the need to avoid space-charge effects and uncontrolled,
12
time-varying reactions in the ion source (which are problematic at high source pressures).
13
14
15
Collector configurations
The ion beams of interest for our analysis include: 12C2H4+ and 13C12CH4+ (for
16
analysis of ethyl fragments), either 12CH3+ and 13CH3+ (our first target when this method
17
was developed) or 12CH2+ and 13CH2+ (our current preferred target) for methyl fragments,
18
and 12C3H8+ and 12C213CH8+ for the propyl species (i.e., the full molecular ion). These ion
19
beams are measured in three separate acquisitions, each having different Faraday cup
20
configurations. These configurations are summarized in Table 1 and briefly described as
21
follows:
22
23
The MAT 253 Ultra has 7 collector positions, 6 of which are movable, three on
either side of the fixed central collector. All collector positions can detect ions using
24
either a Faraday cup, which can be read through an amplifier that varies in gain between
25
107 to 1012, or an electron multiplier. The electron multipliers on the movable collector
26
positions are of the relatively small compact discrete dynode (CDD) design, while that in
27
the central collector position is a larger, conventional secondary electron multiplier
28
(SEM), positioned behind a `retardation lens' (RPQ) for improved abundance sensitivity.
29
Because of the large number of movable collector positions and range of detector
30
sensitivities at each position, it is possible to make three separate measurements of the
31
methyl, ethyl, and propyl fragments simply by changing the magnet current and making
32
minor adjustments of the movable collectors.
33
When 12CH3+ and 13CH3+ are measured, we measure mass 15 on a Faraday cup
34
with a 1012 gain amplifier, on the `L2' collector trolley (Supplemental Figure 1). This
35
collector is positioned so that the high-mass side of the family of mass 15 peaks (i.e.,
36
12
37
are excluded. 13CH3+ is simultaneously measured with the SEM detector in the central
38
collector position, with the magnetic field set such that the low-mass side of the family of
39
mass-16 peaks (i.e., 13CH3+) enters the collector and higher-mass, nearly isobaric species
40
(e.g., 12CH2D+ and 12CH4+) are excluded.
CH3+) enters the collector, and lower-mass ions (mostly 13CH2+ and a trace of 12CHD+)
41
The alternate approach of measuring CH2+ uses a similar cup configuration:
42
12
43
registered through a 1012 ohm amplifier, and 13CH2+ is measured on the low-mass side of
44
the mass 15 peak, using the center collector SEM.
45
46
CH2+ is measured on the high-mass side of the mass 14 peak using the `L3' Faraday cup
When analyzing the ethyl fragment ion, 12C2H4+ is measured on the high-mass
side of the mass 28 peak, collected in a Faraday cup read through a 1012 Ohm amplifier
47
on the `L1' collector trolley (Figure 3), and 13C12CH4+ is measured on the low-mass side
48
of the mass 29 peak, using a Faraday cup registered through a 1012 Ohm amplifier on the
49
central (fixed) collector position.
50
The measurement of propyl fragments involves repositioning one of the movable
51
collectors (L1, or cup 4) after making the ethyl fragment measurement. A 1012 Ω resistor
52
is used on cup 5 to measure mass 45, and a 1011 Ω resistor is used on cup L1 to measure
53
mass 44. We generally do not fully resolve 13C12C2H8+ from 12C3H7D+ at mass 45, which
54
requires a mass resolution close to the limit achieved by the prototype Ultra used in this
55
study (the production version of the Ultra achieves ~2x higher mass resolutions and
56
should be capable of cleaning separating these species). Thus, a minor component,
57
typically a few percent, of each beam will contaminate the other. Our solution to this is to
58
simply combine the signals from these species on the Ultra, and then separately measure
59
the D/13C ratio of the sample on a very high-resolution single collector mass spectrometer,
60
a modified version of the Thermo DFS. The DFS is a high-resolution gas source mass
61
spectrometer, having a reverse geometry and a single collector detector. Eiler et al. 2014,
62
2015 have shown that this instrument can be outfitted with a dual inlet system and used to
63
measure ratios of nearly isobaric isotopologues by untangling the shape of the resulting
64
adjacent peaks (Eiler et al., 2013; Eiler et al., 2014). This measurement, combined with
65
some second constraint (like the sum of 13C+D) typically yields an error of ~0.5-2 ‰ in
66
δD, and no significant increase in error in δ13C beyond that created by that second
67
constraint. For our purposes, we only use this measurement to correct for the contribution
68
of D to the 13C+D signal in the Ultra measurement; this variant of our technique is used
69
for only a small subset of the measurements presented here.
70
71
Analyzer tuning and resolution
At the beginning of each analytical session, the instrument is configured with its
72
high resolution entrance slit and tuned until it achieves a resolution of ~20,000 mass
73
resolving power (M/∆M, 5/95% definition; see (Eiler et al., 2013) for relevant instrument
74
details). The instrument typically retains a stable high-mass resolution for time scales on
75
the order of 1-2 weeks, with minor adjustments to source and analyzer lens potentials
76
made at the beginning of each day.
77
78
79
Signal intensities and counting statistics limits
Given typical source pressures, ion source tuning parameters set to maximize
80
sensitivity, and use of the high-resolution entrance slit, ion currents of the 12CH2+
81
fragment ion are typically on the order of 550 fA, whereas that for the 13C substituted
82
species, 13CH2+, is typically on the order of 5 fA. When the CH3+ fragment ions are
83
instead used to analyze the methyl fragment ion species, intensities are typically 1050 fA
84
for 12CH3+ and 12 fA for 13CH3+. Similarly, typical intensities for the ion beams used to
85
measure the ethyl fragment are 19 pA for 12C2H4+ and 430 fA for 13C12CH4+. 12C3H8+ has
86
a typical intensity of 11 pA, and 13C12C2H8+ has a typical intensity of 360 fA (Table 1).
87
For comparison, a 1 fA ion current represents 6242 ions/sec, and would require ~160
88
seconds of counting to reach a shot-noise-limited precision of 1‰.
89
90
91
92
Background corrections
Given the mass resolutions typically achieved for the instrument and method that
we present, all plausibly significant interferences with the target analyte ion beams are
93
formally resolved (i.e., M/∆M values for all interferences of interest are substantially less
94
than the formal 5/95 % resolution measured when the analyzer is tuned). However, the
95
prototype Ultra used in this study has relatively wide exit slits (i.e., detector apertures) to
96
achieve a `flat-topped peak' condition that facilitates high-precision measurements (the
97
production version of this machine possesses one or more narrower exit slits, obviating
98
some of the issues described below). Therefore, in any case where a target analyte is
99
`sandwiched' between closely adjacent higher- and lower-mass interfering peaks, the
100
analyte ion beam must always be collected along with at least one of those interferences.
101
For this reason, it is essential to our method that the intensities of these co-collected
102
neighboring peaks be measured and used to correct the measured intensity of the analyte
103
plus interference(s). Similarly, corrections must be applied for background contributions
104
from scattered ions and for signals associated with detector dark-noise. Generally, all
105
significant corrections of this kind can be made using a single measurement adjacent to
106
the analyte peak. For example, Supplemental Figure 1 shows a scan of the 16 amu peaks,
107
revealing that 13CH3+ can be delivered to the detector without significant contributions
108
from adjacent 12CH4+, but in this case must include contributions from the other closely
109
adjacent background species, 14NH2+.
110
There are two separate background corrections needed for each measurement. The
111
first is specific to the central collector. The magnet is set to a current typically 200-300
112
BDAC units (or, ~ 0.01 amu) lower than that for the targeted peak, and signal at this
113
position is recorded for two minutes. For the first 60 seconds of this period, gas from the
114
sample reservoir is allowed to flow into the ion source; then, the changeover block is
115
switched so that the standard gas flows into the ion source. After the two minute period is
116
complete, we cull the data to exclude the ~15 seconds of unstable gas pressures
117
associated with switching the changeover valve block, and average over a 30-second
118
period in the middle of each ~1 minute of signal from each gas. After the first
119
background measurement is completed, an isotope ratio acquisition is made, and then a
120
second background measurement is made for the lower-mass species being detected on
121
either cup L1 or L3 (depending on which fragment ion is being measured). If the SEM is
122
in use it is first turned off (since the SEM background on the low mass side of the
123
enriched peak was completed first, and the high mass side of the unenriched peak would
124
swamp the sensor), and then the magnet is set to a current 200-300 BDAC units (~0.01
125
AMU) greater than the magnet setting for the measurement; once more, signal is
126
observed for one minute on each bellows, and averaged for the middle 30 seconds of that
127
time.
128
We find that background intensities are strongly correlated with ion source
129
pressure, presumably because the production and/or ionization efficiency of common
130
background species scales with the pressure of analyte in the ion source. When the
131
pressure within the source is within the range that we specify for our default method,
132
background intensities are approximately stable through time (at least, over the course of
133
several acquisitions). However, if source pressure is significantly higher (i.e., >8x10-8
134
mbar) both of the intensities of monitored ion beams and background intensities
135
measured as described above fluctuate too much for precise (sub-per mil) isotope ratio
136
analysis over seconds to minutes.
137
138
139
Figure Caption:
140
Mass scans illustrating the simultaneous positions of two adjacent collectors during the
141
measurement of 13C/12C ratios of methyl (A) or ethyl (B) fragment ions. In both of these
142
figures, the different colored lines represent mass scans across different collectors. In
143
figure A the shoulder where 13CH2 is measured is clearly wide enough that it is easily
144
resolved from 12CHD. In figure B the shoulder where 13C12CH4 is measured is
145
proportionately narrower but still statistically flat for a great enough range to be
146
measured stably.
147
148
149
150
Eiler J., Blake G. A., Dallas B., Kitchen N., Lloyd M. K. and Sessions A. L. (2014)
Isotopic Anatomies of Organic Compounds. Goldschmidt Conference.
151
152
153
Eiler J., Clog M., Magyar P., Piasecki A., Sessions A. L., Stolper D., Deerberg M.,
Schlueter H.-J. and Schwieters J. (2013) A high-resolution gas-source isotope ratio
mass spectrometer. International Journal of Mass Spectrometry 335, 45–56.
154