Determination of silyl peroxides by ultra-performance liquid chromatography/electrospray ionisation mass spectrometry

Article abstract of DOI:10.1002/rcm.8282

Rationale: Residual initiators in polymers are a concern in the case of products that come directly into contact with the human body or food. Due to low concentrations and difficulties in the sample preparation, highly sensitive and selective methods are required. Methods: A series of bis-silyl- and alkyl-silyl peroxides were analysed by electrospray ionisation mass spectrometry (ESI-MS) on an ultra-performance liquid chromatography/time-of-flight (UPLC/TOF) instrument. Li, Na, K, and NH₄ acetates were used to promote the formation of [M + Me]⁺ ions. The sample preparation involved only dissolution of the polymer sample in 0.1 mL of acetonitrile, followed by precipitation with 1 mL of water. A portion of 0.1–1 μL of the solution was then analysed without further treatment by UPLC/ESI-MS. Results: Limits of detection (LODs) were in the range of 0.06–9 pmol, depending on the peroxide structure. On average, the signal intensity increased with the number of phenyl groups in a peroxide and decreased in the order Na > Li > K > NH₄. Peroxides that did not contain phenyl groups could not be detected. Collision-induced dissociation experiments can be used for structural investigations of alkyl-silyl peroxides. It was possible to detect 2 × 10^?4% (LOD = 7 × 10^?5%) of unreacted Ph₃SiOOt-Bu in the poly(methyl methacrylate) sample. Conclusions: The method is suitable for the analysis of trace peroxide initiators in polymers and for other purposes where LODs in the pmol range are required.

Full text of DOI:10.1002/rcm.8282

UPLC ESI-MS determination of silyl peroxides
Tomasz Krawczyka, Mariusz Zalewskia, Rafał Szołtysika, Anna Korytkowska-Wałachb
aDepartment of Chemical Organic Technology and Petrochemistry, Faculty of Chemistry,
Silesian University of Technology, ul. Krzywoustego 4, 44-100 Gliwice, Poland
bDepartment of Physical Chemistry and Technology of Polymers, Faculty of Chemistry,
Silesian University of Technology, ul. Krzywoustego 4, 44-100 Gliwice, Poland
KEYWORDS: ESI-MS, silyl peroxides, determination, polymerization, initiators
ABSTRACT
Rationale
Residual initiators in polymers are a concern in the case of products that come directly in
contact with human body or food. Due to low concentrations and difficulties in the sample
preparation highly sensitive and selective methods are required.
Methods
A series of bissillyl- and alkyl-silylperoxides were analysed by the electrospray ionisation
mass spectrometry (ESI-MS) on a UPLC-TOF instrument. Li, Na, K, and NH4 acetates were
used to promote the formation of [M+Me]+ ions. The sample preparation involved only
dissolution of a polymer sample in 0.1 mL of acetonitrile, followed by precipitation with 1
mL of water. A portion of 0.1-1 L of the solution was then analysed without further
treatment by UPLC-ESI-MS.
Results
LODs were in the range of 0.06-9 pmol, depending on the peroxide structure. On average, the
signal intensity increased with the number of phenyl groups in a peroxide and decreased in
the order Na>Li>K>NH4. Peroxides that did not contain phenyl groups could not be detected.
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1002/rcm.8282
This article is protected by copyright. All rights reserved.

Collision-induced dissociation experiments can be used for structural investigations of alkyl-
silylperoxides. It was possible to detect 210-4 % (LOD = 710-5 %) of unreacted Ph3SiOOt-
Bu in the poly(methyl acrylate) sample.
Conclusions
The method is suitable for the analysis of trace peroxide initiators in polymers and for other
purposes where LODs in the pmol range are required.
Introduction
Silyl peroxides are a group of compounds containing the Si-OO- linkage either as a
part of a ring or a linear system. Up to four peroxy groups may be connected with a single
silicon atom. Applications and properties of silyl peroxides have been reviewed by various
authors.1–3
The beginning of the research under this class of organometallic compounds can be
dated back to the early 1950s, when they were proposed as polymerisation initiators.4 They
gained a limited, but constant, interest with a few reports appeared every year, roughly half of
them in patent literature.
Styrene or acrylic acid derivatives are among the most frequent monomers
polymerised with silyl peroxides. Thermal4 or Lewis acid-activated free radical
polymerisation5 is possible. Comparing with their carbon counterparts, silyl peroxides are
more stable under high temperatures; therefore, they are effective in the polyethylene
synthesis.6 Silyl peroxides undergo photolysis providing species which can also induce
polymerisation. For example, photodecomposable arylsilyl peroxide in a system with
aluminium compound was used for cationic polymerisation of cyclohexene oxide.7
Among the advantages of the of silyl peroxides as initiators is a low concentration of
the unreacted monomer or the peroxide in the final product. For instance, in the case of
styrene polymerisation the monomer concentration in the polymer can be lowered to 0.03-
0.6%.1
Other reported applications include epoxidation of alkenes,8 Baeyer-Villiger
oxidation9,10 adhesion promoters11 intermediates in the synthesis of antitumour or
antiparasitic drugs.12–17
Detection of silyl peroxides is a challenging task due to the lack of distinct
spectroscopic features of the Si-OO moiety. Low volatility and instability prevents the use of
This article is protected by copyright. All rights reserved.

GC. As they are potent oxidants, indirect methods such as iodometric titration10 or post-
column luminol chemiluminesce18 can be used for this purpose. Nevertheless, such an
approach lacks selectivity and is prone to interferences from hydrogen peroxide or other
organic peroxides.
In the recent years LC-MS gained interest for the detection of various classes of
organic peroxides. In the case of artemisinin,19 TATP20 or lipid peroxides,21 LC-MS is among
the most sensitive methods available. This is due to the progress in soft ionisation techniques,
such as ESI. In most cases alkali metal salts are used as ionisation agents and ions such as
[M+Na]+ or [2M+Na]+ are observed; therefore, individual peroxides can be unequivocally
identified, especially with high resolution instruments.
In this work we attempted to apply UPLC-ESI-MS for the detection of silyl peroxides
aimed at the method that would be suitable for the residual initiator analysis in polymers.
Experimental
Reagents
Cumyl hydroperoxide (CHP), methyl methacrylate (MMA), Trimethylchlorosilan,
phenyldimethylchlorosilan, urea hydrogen peroxide adduct (UHP), and triphenylchlorosilan,
all of reagent grade, and sodium acetate (99%; HPLC grade) were supplied by Acros
Ogranics (Geel, Belgium). Lithium acetate dihydrate (99.999%), acetic acid (99.99%),
LC/MS grade acetonitrile, reagent grade DABCO, tert-butyl hydroperoxide (TBHP),
diphenylmethylchlorosilan, N,N’-bis(trimethylsilyl)urea were purchased from Sigma-Aldrich
(St. Louis, MO). The HPLC-grade acetonitrile was supplied by LiChrosolv (Darmstadt,
Germany). Ammonium acetate (LC/MS grade) was from Fisher Scientific (Fair Lawn, NJ).
Potassium acetate (analytical grade) and benzoyl peroxide (BPO) were purchased from
POCH (Gliwice, Poland). Milli-Q water was used throughout all the experiments.
Methyl methacrylate (MMA) was purified by washing with the NaOH water solution
(5%), then twice with distilled water, and dried over magnesium sulphate. The dried
monomer was finally distilled under reduced pressure (60 mmHg/28-30°C).
Synthesis of standards
Bis(trimethylsilyl)peroxide was synthesised in the reaction of N,N’-
bis(trimethylsilyl)urea with UHP.22
This article is protected by copyright. All rights reserved.

DABCO·2TBHP and DABCO·2CHP were prepared according to Fan et all.23 Briefly,
DABCO (12 mmol) and hexane (30 mL) were added to a 100 mL round-bottomed flask
equipped with a mechanical stirrer. After dissolution of DABCO, 70% solution of TBHP (4.7
mL) or CHP (30 mmol) was added dropwise to the mixture. The resulting precipitate was
filtered at reduced pressure.
In the case of other bis(silylperoxides)23 a 50 mL round-bottomed flask was charged
at 0°C with 12 mmol of triethylamine, 6 mml of UHP, and 20 mL of hexane. The mixture
was continuously mixed with a magnetic stirrer. After 5 min triorganicsilyl chloride (12
mmol) was added dropwise. Precipitation of urea and triethylamine hydrochloride was
immediately observed. After 30 min of stirring at RT, the mixture was filtered and washed
with water several times. After separation, the organic phase was dried with MgSO4 and
evaporated under reduced pressure.
The alkyl-silylperoxides synthesis was based on the published method.23 A 25 mL
round-bottomed flask was charged at 0°C with 2 mmol of DABCO·2TBHP or
DABCO·2CHP, 2 mml of DABCO, and 10 mL of hexane. The mixture was continuously
mixed with a magnetic stirrer. After 5 min, chlorosilane (4 mmol) was added dropwise.
Precipitation of DABCO hydrochloride was immediately observed. After 30 min of stirring at
RT, the mixture was filtered and washed with water several times. The organic phase was
dried by MgSO4 and evaporated under reduced pressure. The structure and purity of all the
1
synthesised peroxides was confirmed by the HPLC-assay, H, and 13C NMR (Agillent 400
MHz).
Instrumentation
Chromatography was carried out using the Waters Acquity UPLC system. The mobile
phase based on acetonitrile (90%) and water (10%) was delivered at 0.2 mL min-1. The BEH
C18 column (1.7 μm; 2.1 × 50 mm) was kept at 30°C. After chromatographic separation, MS
studies were performed with the Waters Xevo G2 QTOF instrument and an ESI source. The
eluate stream was combined with a 5 mM aqueous solution of acetates at a rate of 5 μL min-1.
MS spectra were registered using the ESI+ resolution mode. The conditions for ESI were as
follows: capillary 3 kV, sampling cone 20, extraction cone 4, source temperature 120°C,
desolvation temperature 200°C, cone gas flow 200 L h-1, desolvation gas flow 500 L h-1.
Polymerisation of MMA in bulk
This article is protected by copyright. All rights reserved.

Radical polymerisation of MMA was conducted in two modes: thermal and photo-
induced. TbuPPP (1.5% w/w in respect to monomer) was dissolved in purified MMA (2 mL)
and the obtained solution was degassed by the freeze-pump-thaw cycling prior to
polymerisation. The prepared solution of the initiator in MMA was distributed into a vial and
a quartz tube (0.5 ml each). Thermal polymerisation was conducted in the vial at 60 or 80°C,
for 24 h, in a nitrogen atmosphere. Photopolymerysation was carried out in the quartz tube
using an Osram Ultra-Vitalux® lamp (300 W) for 3 h.
Detection of peroxides in the polymer sample
2 mL of acetonitrile was added to a sample of a freshly prepared polymer and the mixture
was stirred for 2 h. 100 L of the solution was then added to 1 mL of water. The precipitated
polymer was centrifuged and the solution was filtered through a syringe PTFE filter. No
further treatment was necessary before the UPLC-ESI-MS analysis.
RESULTS AND DISSCUSION
The aim of the presented investigation was to develop a method capable of determining silyl
peroxides remaining in the polymer samples. The issue is important if the polymer is
indented for food packaging or for contact with human body24,25.This is a challenging task
mainly due to low concentrations of the initiators in the polymer samples. In this case, ESI-
MS seemed to us as potentially the most sensitive method for this class of compounds, but so
far there have been no reports confirming this assumption if silyl peroxides are considered.
Qualitative ESI-MS analysis
A series of bis(silylperoxides) and silylperoxides shown in Table 1 was selected for this study.
The structures are representative for initiators that are used in industry and in research. The
range of possible structures was limited by the commercial availability of chlorosilanes and
hydroperoxides. The preliminary ESI-MS analysis showed that it is impossible to detect ions
such as [M+H]+, instead intense [M+Me]+ ions were observed in the presence of an alkali
metal or ammonium salts. Relative intensities of ions formed with various alkali metals
depended mainly on the number of phenyl substituents and the symmetry of a molecule. For
This article is protected by copyright. All rights reserved.

compounds containing 1-2 phenyl rings the highest intensities were observed for [M+Na]+
+
ions. Compounds containing 2-4 phenyl rings showed higher affinity towards K+ and NH4 .
The results are summarised in Figure 1.
Because in alkyl-silyl peroxides phenyl substituents could be attached either to the
silicon atom or to the carbon atom, the impact of symmetry of the molecule on the intensities
of ions could be observed. In the case of peroxide derived from tert-butyl hydroperoxide and
diphenylmethylchlorosilane (TBuMPP) the intensities of [M+K]+ and [M+NH4]+ were lower
comparing with CumMMP (cumyl hydrperoxide and phenyldimethylchlorosilane derivative).
A similar difference was visible in the case of TBuPPP and CumMPP. The intensities of
[M+Li]+ ions never exceeded those of [M+Na]+ ions, but their relative intensities increased
with MW of the peroxide (Figure 1).
Collision-induced dissociation (CID) experiments
Depending on the class of the investigated peroxides and to a lesser extent on the
metal ion used, only one fragmentation pattern valid for alkyl-silyl peroxides was identified.
It is shown in Scheme 1.
The first process in Scheme 1 is a neutral loss of either toluene or ethane with the
formation of cation b. It was confirmed based on the average difference between m/z of
cations a and b from Table 2 of 30.048 (RSD=0.004) or 92.071 (RSD=0.018). The expected
difference is 30.047 for ethane and 92.063 for toluene. A loss of H2CO (30.011) or C6H4O
(92.026) was excluded.
Ethane is the favoured product and is exclusively observed for TBuMMP, TBuMPP,
CumMMM, CumMMP, CumMPP. A loss of toluene was observed in the case of TBuPPP
and CumPPP only, where no methyl group is attached to the silicon atom.
Alternative fragmentation of a towards fragment c corresponding to silanones
[R’R”SiO + Me]+ with a neutral loss of either C5H12O, C10H14 or C15H16O depending on the
peroxide was observed when a high collision energy was applied. In-source fragmentation of
parent a allowed to confirm that c is a product of fragmentation of a not b. Fragmentation of
b produced different products of unidentified formulas.
This article is protected by copyright. All rights reserved.

Fragments b and c can be observed only when Na+ or Li+ were used as ionisation
agents. In the case of K+ in the CID experiments only a loss of the metal ion occurred.
The effect of the collision energy on the intensities of each fragment is shown in
Figure 2. It was similar for all the investigated peroxides. A low collision energy of 5 eV
increased the intensity of the parent ion, presumably due to dissociation of the solvent-metal
clusters. A further increase lead to either a loss of the metal, if K+ was considered, or the
formation of the fragment b (Na+ or Li+). Its maximum intensity was observed around 10-12
eV. A further increase of energy lead to the formation of the fragment c. Its intensity was up
to 30% (Na+) or 100% (Li+) of the maximum observed for the parent ion and always the two
heaviest functional groups attached to the silicon atom remained.
All the investigated bis(silyl peroxides) formed no fragment ions during the CID
experiments regardless of the ionisation agent and only a loss of the metal ion from the parent
was observed. All the observed ions are collected in Table 2.
Interestingly, the fragmentation pattern of ammoniated peroxides was different from
the Na, Li or K-doped ones. The parent ion was always [M+NH4]+ of even m/z value and no
[M+H]+ ions were detected. Most of the observed fragments were of odd m/z; therefore, a
loss of nitrogen occurred in most cases. Besides that, only one useful common pattern of
fragmentation was observed. It was noticed that every cumyl-based peroxide (CumMMM,
CumMMP, CumMPP, CumPPP) is able to produce an ion of the observed m/z = 135.07 Da
and the probable formula [C9H11O]+ (calculated m/z = 135.081) (Scheme 2). All parent and
fragmentation ions observed using ammonium acetate are collected in Table 3. The formulas
of fragments a’, b’ and c’ were not unequivocally identified for other compounds.
Quantitative analysis
Absolute LODs were determined by a series of on-column injections of standards of
peroxides made by 10, 100, and 1000-fold dilutions of the stock solution. The injection
volume was also varied between 0.01-2 L in order to obtain a chromatogram where S/N
corresponding to the parent ion was equal to 3. The injected amount as equal to LODs was
presented in Table 4. The procedure was performed in the presence of potassium acetate.
This article is protected by copyright. All rights reserved.

The observed LOD varied between 0.06 and 9 pmol of the peroxide. The lowest limit
was achievable for compounds containing the highest number of phenyl rings and the LODs
increased with the decrease of the number of phenyl groups. In two cases (TBuMMM and
bMMM – only CH3 substituents present) no signal was observed regardless of the injected
amount. Based on the concentration of standards it was calculated that the LODs for those
compounds should be at least 90 pmol, which is beyond practical range. An identical result
was obtained with commercial di-tert-amyl peroxide. The results were surprising, considering
that TATP produce [M+Me]+ ions under the ESI-MS conditions without having aromatic
functionalities. On the other hand, DADP cannot be analysed26, which suggest that the
presence of aromatic rings increases the interaction energy of [M+Me]+, but the steric effects
become important when peroxides such as bPPP or CumPPP are considered. It is visible
when analysing the order of LOD (Table 4). In alkyl-silyl peroxides series each additional
phenyl substituent decreases the LOD roughly by a factor of 10 except TBuMPP and
TBuPPP. In the case of bis(silylperoxides) the LOD decreased 5 times between bMMP and
bMPP, while an addition of another ring on each side of silyl peroxide increases the LOD 20
times when comparing bMPP with bPPP
The feasibility of the quantitative analysis of silyl peroxides was further explored by
registering the calibration curves shown in Figure 3. Essentially, the slopes were proportional
to LODs. The relationship between the injected amount and the [M+K]+ peak intensity was
linear in a narrow range 0-50 pmol for all the compounds and 0-190 pmol for TBuMMP,
TBuMPP, TBuPPP, CumMMM, CumMMP, bPPP, which is well below the amount of the
acetate present. Assuming the 12 s peak width and the amount of potassium acetate infused
(5mM; 5 L min-1) into the eluate stream, it can be calculated that the maximum amount of 5
nmol of peroxide can be ionised. The data indicate that a large excess of acetate is necessary
to achieve a high yield of the ionisation process.
Polymer samples
The feasibility of the method was investigated using poly(methyl methacrylate) samples
obtained with TBuPPP as an initiator under various conditions. For the sake of comparison
BPO was also applied. Because poly(methyl methacrylate) is soluble in acetonitrile, the
sample preparation was straightforward. The polymer was dissolved in acetonitrile and then
precipitated with water. The remaining polymer-free solution was injected on the UPLC
This article is protected by copyright. All rights reserved.

column. The content of the unreacted TBuPPP in the product of thermal polymerisation was
210-4 %. No undecomposed traces of the initiator were detected in the case of the polymer
obtained via photopolymerisation, presumably due to greater efficiency of the latter. This is
consistent with literature reports that decomposition of some silyl peroxides is accelerated by
UV light.27 Based on the data in Table 4 and the amounts of polymer and solution used it can
be calculated that the LOD in polymer sample was 710-5 %, which is in a satisfactory range
for the intended purposes. For other peroxides it was in the range of
310-3 % to 210-5 % (calculated from Table 4). The content of BPO detected in the
polymers was 1510-3 % and 4410-3 % for the thermal polymerisation samples at 80 and
60°C, respectively. The photopolymerisation sample contained 5310-3 % of the unreacted
BPO, which is much higher comparing with TBuPPP.
CONCLUSIONS
Silyl peroxides can be detected with an ESI-MS instrument at LOD of 9-0.06 pmol
depending on the structure. Na and K acetates were the best doping agents. The sensitivity
was low enough to allow the detection of silyl peroxides at of 310-3 % to 210-5 % in the
actual polymer with straightforward sample preparation. The proposed method could be
applied for studies in polymer sciences where traces of peroxides as initiators need to be
quantified.
Acknowledgements
The publication has been supported by Rector’s grant in the scientific research and
development area. Silesian University of Technology Grant number 04/050/RGJ17/0048.
The authors have declared no conflict of interest.
References
1
Y. A. Aleksandrov. Preparation, properties and applications of organosilicon
This article is protected by copyright. All rights reserved.

peroxides. J. Organomet. Chem. 1982, 238, 1.
2
J. C. Hill. Advances in Silicon Chemistry, Vol. 3 Edited by Gerald L. Larson (Huls
America, Inc., New Jersey). JAI Press:ꢀ Greenwich, CT. 1996. ix + 279 pp. $109.50.
ISBN 1-55938-831-5. J. Am. Chem. Soc. 1997, 119, 8582.
A. G. Davies. Organosilicon peroxides: radicals and rearrangements. Tetrahedron
2007, 63, 10385.
3
4
N. V. Yablokova, Y. A. Aleksandrov, O. M. Titova. Styrene polymerization on
mineral filler surfaces, modified with organosilicon peroxides. Polym. Sci. U.S.S.R.
1986, 28, 2122.
5
Y. D. Semchikov, N. A. Kopylova, N. V. Yablokova, L. N. Nistratova. Initiating
radical polymerization of vinyl monomers by silicon peroxides. Eur. Polym. J. 1986,
22, 569.
6
7
Y. Fan, R. Greggshaw. Coupling solid substrates using silyl peroxide compounds. US
Grant: US3856606A. 1974.
S. Hayase, Y. Onishi, S. Suzuki, M. Wada. Photopolymerization of Cyclohexene
Oxide by the Use of Photodecomposable Silyl Peroxide. Aluminum Complex/Arylsilyl
Peroxide Catalyst. Macromolecules 1986, 19, 968.
8
C. M. Estévez, O. Dmitrenko, J. E. Winter, R. D. Bach. Reactivity of alkyl versus silyl
peroxides. The consequences of 1,2-silicon bridging on the epoxidation of alkenes
with silyl hydroperoxides and bis(trialkylsilyl)peroxides. J. Org. Chem. 2000, 65, 8629.
M. Suzuki, H. Takada, R. Noyori. Bis(trimethylsilyl) Peroxide for the Baeyer-Villiger
Type Oxidation. J. Org. Chem. 1982, 47, 902.
9
10
S. Baj, A. Chrobok, R. Słupska. The Baeyer–Villiger oxidation of ketones with
bis(trimethylsilyl) peroxide in the presence of ionic liquids as the solvent and catalyst.
Green Chem. 2009, 11, 279.
11
12
13
14
Y. L. Fan. Silyl Peroxide-Some Chemical and Physical Effects on Adhesion
Promotion. J. Adhes. 1972, 4, 261.
B. Barnych, J. M. Vatèle. Exploratory studies toward the synthesis of the
peroxylactone unit of plakortolides. Tetrahedron 2012, 68, 3717.
A. Ramirez, K. A. Woerpel. Synthesis of 1,2-dioxolanes by annulation reactions of
peroxycarbenium ions with alkenes. Org. Lett. 2005, 7, 4617.
E. Radzikowska, J. Baraniak. Synthesis of PS/PO-chimeric oligonucleotides using
mixed oxathiaphospholane and phosphoramidite chemistry. Org. Biomol. Chem. 2015,
13, 269.
This article is protected by copyright. All rights reserved.

15
16
K. Igawa, K. Sakita, M. Murakami, K. Tomooka. Partial oxidation of alkenylsilanes
with ozone: A novel stereoselective approach to the diol and triol derivatives.
Synthesis (Stuttg). 2008, 1641.
Z.-J. Xu, S. Wittlin, Y. Wu. Probing the Peroxycarbenium 3+2 Cycloaddition
Reactions with 1,2-Disubstituted Ethylenes: Results and Insights. Chem. - A Eur. J.
2017, 23, 2031.
17
18
Z. J. Xu, D. X. Tan, Y. Wu. Synthesis and Structural Reassignment of Plakinidone.
Org. Lett. 2015, 17, 5092.
S. Baj, R. Słupska, T. Krawczyk. Application of response surface methodology
(RSM) to the optimization of a post-column luminol chemiluminescence analysis of
silyl peroxides. Talanta 2013, 103, 172.
19
20
21
M. Wang, C. Park, Q. Wu, J. E. Simon. Analysis of artemisinin in Artemisia annua L.
by LC-MS with selected ion monitoring. J. Agric. Food Chem. 2005, 53, 7010.
A. Stambouli, A. El Bouri, T. Bouayoun, M. A. Bellimam. Headspace-GC/MS
detection of TATP traces in post-explosion debris. Forensic Sci. Int. 2004, 146, 191.
I. Milic, M. Fedorova, Derivatization and detection of small aliphatic and lipid-bound
carbonylated lipid peroxidation products by ESI-MS, in Advanced Protocols in
Oxidative Stress III. 2014, pp. 3–20.
22
23
24
W. P. Jackson. A Simple Preparation of Bis(trimethylsilyl) Peroxide. Synlett 1990,
1990, 536.
Y. L. Fan, R. G. Shaw. Amine-Hydroperoxide Adducts. Use in Synthesis of Silyl
AlkyI Peroxides. J. Org. Chem. 1973, 38, 2410.
P. J. Fordham, J. W. Gramshaw, L. Castle. Analysis for organic residues from aids to
polymerization used to make plastics intended for food contact. Food Addit. Contam.
2001, 18, 461.
25
26
27
Euroean Commission. Commission Regulation (EU) No 10/2011 of 14 January 2011
on Plastic materials and articles intended to come into contact with food. Off. J. Eur.
Union 2011, 15, 12.
M. E. Sigman, C. D. Clark, T. Caiano, R. Mullen. Analysis of triacetone triperoxide
(TATP) and TATP synthetic intermediates by electrospray ionization mass
spectrometry. Rapid Commun. Mass Spectrom. 2008, 22, 84.
R. L. Dannley, G. Jalics. The Decomposition of Silyl Hydroperoxides. J. Org. Chem.
1965, 30, 3848.
This article is protected by copyright. All rights reserved.

Table 1 Structures of organic silylperoxides and bis(silylperoxides) used in this study
Substituent
R’
Symbol
Structure
R’’’
R
R’’
-CH3
-C6H5
-C6H5
-C6H5
-CH3
-C6H5
-C6H5
-C6H5
-CH3
-C6H5
-C6H5
-C6H5
-CH3
-CH3
-CH3
-CH3
-C6H5
-CH3
-CH3
-CH3
-C6H5
-CH3
-CH3
-CH3
-C6H5
TBuMMM*
TBuMMP
TBuMPP
TBuPPP
CumMMM
CumMMP
CumMPP
CumPPP
bMMM*
bMMP
-CH3
-CH3
-C6H5
-C6H5
-CH3
-CH3
-C6H5
-C6H5
-C6H5
-CH3
-CH3
-
-C6H5
-C6H5
bMPP
bPPP
*not detected in the actual samples
This article is protected by copyright. All rights reserved.

Table 2 Parent and fragment ion m/z observed during the ESI-MS analysis of silyl peroxides in
the presence of alkali metal acetates
Metal
K+ [Da]
Na+ [Da]
Li+ [Da]
Peroxide
TBuMMM
TBuMMP
TBuMPP
TBuPPP
CumMMM
CumMMP
CumMPP
CumPPP
bMMM
a
b
c
-
-
-
-
-
-
-
-
-
-
-
-
a
b
c
a
b
c
-
-
-
-
-
-
-
-
263.09
325.09
387.09
263.08
325.09
387.09
449.11
-
-
-
-
-
-
-
-
-
-
-
-
247.10 217.06 159.03
309.11 279.06 221.03
371.12 279.06 221.03
247.10 217.06 97.00
309.11 279.06 159.03
371.12 341.08 221.03
433.13 341.08 221.03
231.14 201.09 143.04
293.14 263.09 205.05
355.19 263.09 205.05
231.14 201.08 81.00
293.14 263.09 143.04
355.15 325.10 205.05
417.18 325.10 205.05
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
bMMP
341.07
465.09
589.14
325.09
449.12
573.16
309.12
433.16
557.19
bMPP
bPPP
This article is protected by copyright. All rights reserved.

Table 3 Parent and fragment ion m/z observed during the
ESI-MS analysis of silyl peroxides in the presence of
ammonium acetate
a
b’
c’
d’
TBuMMP
TBuMPP
TBuPPP
CumMMM
CumMMP
CumMPP
CumPPP
bMMP
242.15
151.06
134.06
108.04
304,16
366.17
-
213,00
171,00
215.04
135.06
135.07
135.07
135.07
209.03
271.01
-
153,00
275.07
-
-
107.04
304.16
366.17
428.20
320.14
444.13
568.20
-
107.04
-
-
-
-
-
-
-
225.06
331.05
395.08
bMPP
bPPP
This article is protected by copyright. All rights reserved.

Table 4 Limits of detection (LOD) of
silylperoxides achieved with the UPLC/ESI-
MS detection and K(CH3COO) as an ion-
pairing reagent
LOD
(pmol)
Compound
m/z
201
263
325
387
263
325
387
449
217
341
465
589
242
270
>90*
TBuMMM
8.7
0.25
0.24
8.6
TBuMMP
TBuMPP
TBuPPP
CumMMM
0.68
0.06
0.08
>90*
0.65
0.14
3.3
CumMMP
CumMPP
CumPPP
bMMM
bMMP
bMPP
bPPP
1.6
BPO
CPO
0.13
*not detected in the actual samples
This article is protected by copyright. All rights reserved.

Figure 1 Relationship between the structure of peroxides and the relative intensity of the
observed [M+Me]+ ions. Na = 1
This article is protected by copyright. All rights reserved.

Figure 2 Influence of the collision energy on the intensity of the fragment ions of alkyl-silyl
peroxides. Addition of K – A; Na, Li – B
This article is protected by copyright. All rights reserved.

Figure 3 Effect of the injected amount on the absolute signal intensities of [M+K]+ ions of
silyl peroxides under the ESI conditions. 90% acetonitrile, 10% water, 5 mL of 5 mM min-1
of CH3COOK. Error bars represent standard uncertainties n = 3
This article is protected by copyright. All rights reserved.

Scheme 1 Fragmentation pattern of alkyl-silylperoxides
This article is protected by copyright. All rights reserved.

Scheme 2 Common fragmentation pattern of cumyl-silyl peroxides in the presence of
ammonium acetate
This article is protected by copyright. All rights reserved.