09

Mobile Phase Additives for LC-MS. Part 3: The Neutral Salts ........ This is the third article in a
five part series on mobile phase additives for LC-MS to appear in each issue of Analytix in 2006
By Joachim Emmert, Applications Development R&D, Fluka / Riedel-de Haën … [email protected]
Alexander Leitner, Department of Analytical Chemistry, University of Vienna … [email protected]

Although organic acids are the most common mobile This short article will discuss the characteristics, benefits
phase additive for HPLC separations that employ MS and practical use of the ammonium salts of acetic and
detection, it may be necessary under certain formic acid as LC-MS mobile phase additives. All
circumstances to use more neutral conditions, either analytical conditions and test compounds were the
because the analytes are sensitive to acids or do not same as described in part 1 of this series [1], except the
exhibit optimal resolution at low pH. When acids are concentration of raffinose, which was 100 ng/mL in this
not suitable, volatile salts, like ammonium formate or study. Additionally, a four peptide mixture of bradykinin
ammonium acetate, may be the additives of choice analogues was used in one experiment. The salts were
(Table 1). However, compared to organic acids their use dissolved in the aqueous part of the mobile phase at a

Chromatography
is much more complex. One issue is the limited solubility concentration of 0.1% w/v. The organic part of the
of the salts in organic solvents; another issue is the mobile phase was used either without any additive or as
changing pH value during a gradient. On the other ready-to-use LC-MS CHROMASOLV® blends also
hand, the mildly acidic pH provided by the salts permits containing 0.1% w/v additive (Table 2).
both positive and negative ion mode detection.

Table 1 ......... List of Sigma-Aldrich LC-MS additives

Cat. No. Brand Description* Package Size Packaging
40967 Fluka Trifluoroacetic acid, puriss p.a., eluent additive for LC-MS 50 mL HDPE bottle
40967 Fluka Trifluoroacetic acid, puriss p.a., eluent additive for LC-MS 10 x 1 mL Glass ampuls
56302 Fluka Formic acid, puriss p.a., eluent additive for LC-MS 50 mL HDPE bottle
49199 Fluka Acetic acid, puriss p.a., eluent additive for LC-MS 50 mL HDPE bottle
49916 Fluka Propionic acid, puriss p.a., eluent additive for LC-MS 50 mL HDPE bottle
55674 Fluka Ammonium formate, puriss p.a., eluent additive for LC-MS 50 g HDPE bottle
49638 Fluka Ammonium acetate, puriss p.a., eluent additive for LC-MS 50 g HDPE bottle
61333 Fluka Sodium citrate tribasic dihydrate, puriss p.a., eluent additive for LC-MS 50 g HDPE bottle
40867 Fluka Ammonium bicarbonate, puriss p.a., eluent additive for LC-MS 50 g HDPE bottle
44273 Fluka Ammonium hydroxide solution 25%, puriss p.a., eluent additive for LC-MS 100 mL HDPE bottle
65897 Fluka Triethylamine, puriss p.a., eluent additive for LC-MS 50 mL HDPE bottle
* ”puriss” quality grade is defined as >98.5% assay, <0.1% ash, and specification n + 0.001, d + 0.001 with no extraneous color and a homogeneous
appearance. “p.a.” or pro analysi denotes a product with guaranteed trace impurity levels and/or suitability for the indicated analytical application.

Table 2 ......... Selection of LC-MS CHROMASOLV® blends

Cat. No. Brand Description Package Size Packaging
34674 Riedel-de Haën Water with 0.1% ammonium acetate LC-MS CHROMASOLV® 2.5 L amber bottle
34670 Riedel-de Haën Methanol with 0.1% ammonium acetate LC-MS CHROMASOLV® 2.5 L amber bottle
34669 Riedel-de Haën Acetonitrile with 0.1% ammonium acetate LC-MS CHROMASOLV® 2.5 L amber bottle
34668 Riedel-de Haën Acetonitrile with 0.1% formic acid LC-MS CHROMASOLV ®
2.5 L amber bottle

www.sigma-aldrich.com/lc-ms-additives
 10

The main issue when using ammonium acetate or Besides affecting the apparent pH, using buffered
ammonium formate as additives is their solubility, which mobile phase components has a significant impact on
is very good in water, sufficient in methanol to obtain a the separation and ionization of the test compounds in
concentration of 0.1% w/v, but not in acetonitrile. This this study. Under conditions C (Figure 2) and conditions
is a problem since acetonitrile is the organic solvent of A (Figure 3), reserpine (blue peak) is shifted in retention
choice for most separations. The effect is shown in time, and both reserpine and propazine (dark green)
Figure 1. When using pure acetonitrile as the organic exhibit different degrees of ionization. The effect is even
part in gradient elution against 0.1% ammonium more pronounced on the four bradykinin analogues
acetate in water, the apparent pH will rise and influence (Figure 4). Resolution was greatest using the buffered
the separation, worsening it in most cases (curve A). The conditions C (upper trace). However, unfortunately it
same is true when running a gradient with methanol also had a higher tendency to form sodium adducts when
containing 0.1% ammonium acetate in both solvents using ion trap instruments compared to triple quads [2].
(curve B). To address the solubility issue, Sigma-Aldrich
has developed a special blend (Cat. No. 34669, patent Similar observations are made for ammonium formate.
pending), which contains 0.1% w/v of ammonium Table 3 lists the changes in pH when using 0.1%
Chromatography

acetate in acetonitrile stabilized with acid. This acid-

stabilization has three desirable effects: the salt is kept Figure 1 .......... Apparent pH curves using buffered and
in solution, the blend is stable against decomposition unbuffered blends

and the system is buffered. It also keeps the pH in the

pH gradient with different ammonium acetate blends
mildly acid range when using both the aqueous and pH 8.50

organic components as buffered blends (curve C) and A — A: H2O / 0.1% NH4Ac B: CH3CN 100.0

8.00
B — A: H2O / 0.1% NH4Ac B: CH3OH / 0.1% NH4Ac
when using the acetonitrile blend not as intended, but C — A: H2O / 0.1% NH4Ac B: CH3CN / 0.1% NH4Ac
D — A: H2O B: CH3CN / 0.1% NH4Ac
with pure water as the aqueous solvent (curve D). 7.50

7.00
A
Figure 2 .......... EIC (positive ion mode) of test compounds with ammonium acetate as additive B
6.50
in both aqueous and organic components (conditions C in Figure 1)
Red trace is raffinose, green is bradykinin, dark red is digoxin, blue is reserpine, and dark green is propazine. 6.00

5.50
Intens.
x106
5.00

Gradient conditions C: C
5
A: H2O containing 0.1% w/v NH4Ac 4.50
B: CH3CN containing 0.1% w/v NH4Ac
4
4.00 D
ACN: 0.0+/+ 0.0

3 Flow: 0.30 mL/min

3.50
7.0 20.0 30.0 40.0 50.0 64.0
2 min

Figure 4 .......... EIC (positive ion mode) of bradykinin analogues

0
0 5 10 15 20 25
with both components containing buffered ammonium acetate
Time [min]
EIC 527.1 ±All EIC 530.7 ±All MS
EIC 609.3 ±All EIC 839.5 ±All MS
EIC 230 ±All
(upper trace) and with 0.1% ammonium acetate / acetonitrile, not
buffered (lower) (conditions C and A in Figure 1)
Figure 3 .......... EIC (positive ion mode) of test compounds with ammonium acetate in the 1 = bradykinin 1-6, 2 = Lys-Ala3-bradykinin, 3 = bradykinin,

aqueous component only (conditions A in Figure 1) 4 = des-Arg1-bradykinin, 5 = impurity

Red trace is raffinose, green is bradykinin, dark red is digoxin, blue is reserpine, and dark green is propazine.
Intens. 4 31010502.D:TIC ±AII
x107
Intens. 3
4 Gradient conditions C:
x106
1 Aqueous and organic components
6 Gradient conditions A: 5 both buffered (0.1% w/v
A: H2O containing 0.1% w/v NH4Ac 3 2 ammonium acetate, stabilized)
5 B: CH3CN
2
4
1
3

x108 28010502.D:TIC ±AII

1.2
4
2
Gradient conditions A:
1.0 1
Aqueous component buffered (0.1%
1
w/v ammonium acetate, stabilized),
0.8 3
acetonitrile (not buffered)
0
0 5 10 15 20 25 Time [min] 0.6
EIC 527.1 ±All EIC 530.7 ±All MS
EIC 609.3 ±All EIC 839.5 ±All MS
EIC 230 ±All
0.4
2
0.2

0.0
5 10 15 20 25 30 35 Time (min)

www.sigma-aldrich.com/lc-ms-additives
 11

Figure 5 .......... EIC (positive ion mode) of test compounds with ammonium formate in the ammonium formate in gradients with either pure
aqueous component only acetonitrile or with acetonitrile spiked with 0.1% formic
Red trace is raffinose, green is bradykinin, dark red is digoxin, blue is reserpine and dark green is propazine.
acid (FA). This latter combination functions as kind of
Intens. “on the fly buffering,” which significantly affects the
x107
separation and ionization, although the apparent pH
3 Gradient:
A: H2O containing 0.1% w/v NH4 formate
differences are not that dramatic.
B: CH3CN

2 Table 3 ......... pH change during gradient of acetonitrile against

ammonium formate (aqueous component: 0.1% w/v ammonium
formate in water)
1
% aqu. % CH3CN pH CH3CN pH CH3CN / 0.1% FA
100 0 6.3 6.3
0 5 10 15 20 25 Time [min]
EIC 527.1 ±All EIC 530.7 ±All
50 50 6.9 6.7
EIC 609.3 ±All EIC 651.4 ±All
EIC 230 ±All
10 90 7.5 7.1

Figure 6 .......... EIC (negative ion mode) of test compounds with ammonium formate in the
aqueous component only Red trace is raffinose, green is bradykinin, dark red is digoxin, blue is reserpine; Figure 5 shows the test mix separation using a gradient
propazine is not detected in neg. ion mode. between 0.1% ammonium formate and pure
acetonitrile. Under these conditions detection in
Intens.
x106 negative ion mode is also possible, which often results
1.5
Gradient: in a more specific and less noisy signal (Figure 6). In
A: H2O containing 0.1% w/v NH4 formate
Figure 7 perfect resolution is achieved when using

Chromatography
B: CH3CN

1.0 water with 0.1% w/v ammonium formate and

acetonitrile with 0.1% v/v formic acid.
0.5

An interesting observation worthy of discussion are the

mass spectra of the test components obtained in
0.0
0
EIC 503.1 ±All
5 10 15
EIC 1058.5 ±All
20 25 Time [min]
negative ion mode (Figure 8). The normal molecular ion
EIC 825.4 ±All
is [M-H]-, 503.2 for raffinose, 779.4 for digoxin, 607.3
EIC 607.3 ±All
EIC 228.1 ±All

for reserpine and 1058.6 for bradykinin. In this case only

Figure 7 .......... EIC (positive ion mode) of test compounds with ammonium formate in the the singly charged molecular ion is observed for the
aqueous component and formic acid in the organic component
peptide bradykinin, contrary to positive ion mode,
Red trace is raffinose, green is bradykinin, dark red is digoxin, blue is reserpine and dark green is propazine.
where the doubly charged ion is dominant. For the
Intens.
x107
other test compounds addition of one formate anion,
1.50
[M+45]-, is also observed.
Gradient:
1.25
A: H2O containing 0.1% w/v NH4 formate
B: CH3CN containing 0.1% v/v formic acid
1.00 In conclusion, the neutral volatile salts, ammonium
0.75
acetate and ammonium formate, offer a much broader
influence on analyte separation and ionization than do
0.50
the acids. Their use, of course, is dictated by the
0.25
particular LC-MS separation objectives or problems
0.00
0 5 10 15 20 25 Time [min]
being addressed. Any limitations to their solubility may
EIC 527.1 ±All
EIC 609.3 ±All
EIC 230 ±All
EIC 530.7 ±All
EIC 803.4 ±All actually turn into the possibility of doing the separation
or detection in a really unusual way.
Figure 8 .......... Mass spectra of test compounds in negative ion mode shown in Figure 6
References:
Intens.
x105
-MS. 1.5 min (#35) [1] “Mobile Phase Additives for LC-MS. Part 1: Acids – The Most
2 617.1 Common Choice,” Analytix 2006/2, 8-9.
1 503.1 549.1
1007.2 (See also: “Mobile Phase Additives for LC-MS. Part 2: How to
0
x104 -MS. 16.3 min (#379)

6
Overcome Suppression Effects of TFA,” Analytix 2006/3, 16-17.
1058.5
4 Both downloadable from:
2
373.0 416.9 487.2
0
x105 -MS. 17.6 min (#47)
http://www.sigma-aldrich.com/analytix
6
4 825.4 893.4 [2] “Influence of solvent additive composition on chromatographic
2
779.4
separation and sodium adduct formation of peptides in HPLC-
0
x104
7.5
-MS. 17.6 min (#47)
ESI-MS”, Poster at HPLC 2006 San Francisco, June 2006; will
607.2
5.0 653.2 appear in J. Chromatogr. A, symposium issue.
2.5
277.0 480.8 544.9 721.1 799.1 830.5 908.5
0.0
0 200 400 600 800 1000 m/z

www.sigma-aldrich.com/lc-ms-additives