Volatiles Evolved From Decaying Human

www.elsevier.com/locate/forsciint

Forensic Science International 153 (2005) 147–155

A study of volatile organic compounds evolved from the

decaying human body

M. Statheropoulosa, C. Spiliopouloub, A. Agapioua,*

aSchool of Chemical Engineering, National Technical University of Athens (NTUA),

Sector I, 9 Iroon Polytechniou Str., 157 73 Athens, GreecebDepartment of Forensic Medicine and Toxicology, Medical School, University of Athens, 115 27 Goudi, Greece

Received 1 October 2003; accepted 13 August 2004

Available online 25 November 2004

Abstract

Two men were found dead near the island of Samos, Greece, in the Mediterranean sea. The estimated time of death for both

victims was 3–4 weeks. Autopsy revealed no remarkable external injuries or acute poisoning. The exact cause of death remained

unclear because the bodies had advanced decomposition. Volatile organic compounds (VOCs) evolved from these two corpses

were determined by thermal desorption/gas chromatography/mass spectrometry analysis (TD/GC/MS). Over 80 substances

have been identified and quantified. The most prominent among them were dimethyl disulfide (13.39 nmol/L), toluene

(10.11 nmol/L), hexane (5.58 nmol/L), benzene 1,2,4-trimethyl (4.04 nmol/L), 2-propanone (3.84 nmol/L), 3-pentanone

(3.59 nmol/L). Qualitative and quantitative differences among the evolved VOCs and CO2 mean concentration values might

indicate different rates of decomposition between the two bodies. The study of the evolved VOCs appears to be a promising

adjunct to the forensic pathologist as they may offer important information which can be used in his final evaluation.

# 2004 Elsevier Ireland Ltd. All rights reserved.

Keywords: Evolved VOCs; TD/GC/MS analysis; Corpse; Volatiles; Odors; Cadaver

1. Introduction

Over the last decade interest has increased regarding the

identification of VOCs for medical [1,2], toxicological [3,4]

and environmental applications [5]. Forensic science uses

headspace analysis for volatile substances in blood and

organ specimen (i.e. brain, lung, liver, kidney, skeletal

muscle); this approach has been a well-established method

[6,7]. The odor of putrefaction is characteristic and familiar

to the front line experts such as police investigators, forensic

pathologists, anthropologists, entomologists, crime scene

technicians and other medical and non-medical profes-

* Corresponding author. Tel.: +30 210 7723109;

fax: +30 210 7723188.

E-mail address: [email protected] (A. Agapiou).

0379-0738/$ – see front matter # 2004 Elsevier Ireland Ltd. All rights r

doi:10.1016/j.forsciint.2004.08.015

sionals. The chemical analysis of the malodorous com-

pounds has not attracted much of research interest. It has

been referred that investigators do not attempt to identify and

quantitate these compounds because routine methods cannot

detect them at their low concentrations [8]. However, these

odors are readily sensed by various insects with their sensi-

tive olfaction. Forensic entomology has been used for

determining time interval since death [9,10]. Furthermore,

it is speculated that the odor of putrefaction comes primarily

from the sulfur compounds and various inorganic gases that

are produced in the bowel.

The breakdown of soft tissues of the body during putre-

faction is aided by microorganisms such as bacteria, fungi

and protozoa. This results in the production of gases, liquids

and simple molecules. Various gases (CO2, H2S, CH4, NH3,

SO2, H2) produced in the process distend the tissues [11].

eserved.

M. Statheropoulos et al. / Forensic Science International 153 (2005) 147–155148

Along with these, a variety of volatile organic compounds

(VOCs) are liberated. These volatile substances are inter-

mediate products of decomposition while the large biologi-

cal macromolecules such as proteins, nucleic acids,

carbohydrates and lipids are broken down into their building

block components (C, H, O, N, P and S). Carbohydrates are

broken down mainly to oxygenated compounds (alcohols,

aldehydes, ketones, acids, esters and ethers) and proteins to

nitrogen, phosphorus and sulfur compounds. Nucleic acids

produce nitrogen and phosphorus compounds. Lipids give

rise to hydrocarbons, nitrogen, phosphorus and oxygenated

compounds.

Many factors affect the decomposition rate of a corpse.

Above ground decomposition is typically faster than burials

because the corpse is exposed to varying environmental

factors (i.e. temperature, humidity, air currents), insects

and carnivores [12]. Burials in many forms slow decom-

position by decreasing gas diffusion, limiting both macro

and microorganisms and increasing the amount of carbon

dioxide leading to anaerobic conditions [13].

Sampling volatile organic compounds onto sorbent tubes

and their subsequent analysis by TD/GC/MS is an efficient

method for qualitative and semi-quantitative determination of

VOCs. The aim of the work presented here is the determina-

tion of the type and amount of VOCs present at a single time

point during the decay process of the human body. Consider-

ing the given set of circumstances of each death case, VOCs

may offer valuable information which could prove useful in

estimating the postmortem interval. Moreover, this data could

also help uncover the etiology of death.

2. Case profiles and autopsy findings

Two human males were found dead in December 2002

near the island of Samos in Greece. The port authorities

enclosed each cadaver in a plastic body bag which was then

placed in a metal case and air sealed with silicone in a

wooden box. Bad weather travel conditions prevailed at the

time. So the bodies were shipped 4 days later to the medical

school mortuary in Athens for medico-legal examination.

2.1. Case study 1

The cadaver belongs to an unknown man, 183 cm tall,

slim and aged about 40 years (between 34 and 44 years old).

Although the cause of death remains uncertain, the victim

couldhavedrownedbecause a largeamountoffluid (seawater)

was found in his chest cavity. Autopsy revealed that there

were neither remarkable external injuries nor acute poisoning.

Most internal organs showed softening by putrefaction.

2.2. Case study 2

The cadaver belongs to an unknown man, 183 cm tall,

robust, and aged about 30 years (between 26 and 36 years

old). The cause of death could not be determined because

decomposition had progressed. Although the external

appearance of the corpse was better preserved in relation

to the first case, putrefaction was more advanced. Decay was

prominent in the internal organs.

3. Experimental

3.1. Chemicals

Five standard mixtures (C = 50 mmol/L) were purchased

from ITChem Hellas. In order to determine the relative

response factors (RRF) for the applied semi-quantitative

method, four different methanolic standard mixtures of 12

substances each (H/C, O-atoms, heteroatoms and aromatic

substances) were used. Furthermore, a standard solution of

deuterated chlorobenzene-d5 99% in methanol was used as

internal standard. Chlorobenzene-d5 is a better standard for

high molecular weight and low volatility compounds. How-

ever, it was chosen for the applied full screening semi-

quantitative method as it complies to the general rules of

EPA TO-17 for gas standards [14]. Furthermore, its elution

occurs in the middle of the chromatographic analysis (mean

retention time R̄t = 28.6).

3.2. Sampling tube preparation

Three layer sampling sorbent glass tubes of 4 mm i.d. and

115 mm length were used for VOC measurement (TD-300,

Alltech Associates). These consisted of 300 mg carbograph

2, 200 mg carbograph 1 and 125 mg carbosieve S-III. The

tubes were conditioned for 2 h at 300 8C at a flow of 150 mL/

min of He to minimize background effects. The above

conditioning procedure was verified for success by blank

measurements. The sampling tubes were sealed with brass

nuts, caps and PTFE ferrules and kept at 4 8C until the day of

sampling. Before sampling, 1 mL of the above methanolic

standard mixture of chlorobenzene-d5 was added by means

of microsyringe to the glass wool supporting the sorbent

material.

3.3. VOCs, CO2 and NH3 sampling

Each wooden box (2.20 m � 0.52 m � 0.60 m) was

opened and the metal case was layed open. A small hole

was made at the body bag at the upper chest level of each

cadaver and a Teflon tube (0.6 cm i.d.) was inserted. The

sorbent tube was placed at the end of the Teflon tube. Five

liters of air (‘‘headspace’’) were drawn from the body bags

through the sampling tube by a sampling pump (VSS-1, A.P.

Buck, USA) at a flow rate of 200 mL/min. A reference

sample was also taken for comparison outside the cadaver

bag at a close distance.

A second, smaller Teflon tube (0.1 cm i.d.) was inserted

through the body bag of the cadaver in order to monitor

M. Statheropoulos et al. / Forensic Science International 153 (2005) 147–155 149

carbon dioxide concentrations during VOCs sampling. A

portable IR gas analyzer (Anagas CD 98, Environmental

Instruments, England) was used with a PTFE filter provided

by the instrument manufacturer. The flow rate of the instru-

ment with the front filter was 120 mL/min.

A third Teflon tube was finally inserted through the body

bag of each cadaver to monitor ammonia concentration as

well. A portable ammonia gas sensor without internal pump

(Toximeter TX 2000, Oldham, France) was used. The

ambient temperature and relative humidity during sampling

were measured by a portable thermometer-hygrometer

(model H 270, Dostmann Electronics, Germany).

Actually three variations of sampling were employed for

the qualitative analysis of VOCs. The first method was

described above. In the second method the VOCs were first

placed in a 2 L flexible Tedlar bag and then transferred to the

sorbent tube. This method was chosen because it is widely

used for expired air analysis where preconcentration of the

sample is necessary. In order to avoid sample pass through

the pump, an in house vacuum chamber was used. The

Tedlar bag was filled directly using the negative pressure

provided by the pump [15]. The number of VOCs identified

with this method was slightly lower than the previous one

and quantitative differences among VOCs appeared, as

lower volume of air was sampled. These findings were

expected since this method involved multiple analytical

steps where leaks may occur and VOCs may also be retained

in the bag. In the third method, a sorbent tube was inserted

directly in each cadaver bag and air was allowed to diffuse

into it without a pump. This experiment showed an even

lower number of VOCs and only the higher peaks could be

identified. For the qualitative analysis, the results of the three

sampling methods were taken into consideration. For semi-

quantitation purposes, the first method was used since it

involves fewer steps and experimental conditions were

controlled.

In order to examine if VOCs originate from the plastic

packing material of the body bag, a solid phase micro

extraction (SPME) experiment was carried out. An empty

body bag (1.87 m � 0.87 m, Adamedical, Hellas) was filled

with synthetic air and left for 1 day at ambient temperature.

The SPME (85 mm carboxen/polydimethylsiloxane on a

Stableflex fiber, Supelco) was inserted through the body

bag and the fiber was exposed for 30 min, equal to the actual

time of sampling into the sorbent tubes.

3.4. Chromatographic analysis

Sorbent tubes were thermally desorbed to an HP 5890/

5972 GC/MS system using an in house made thermal

desorption unit (TDU). This unit is designed to stand on

top of the gas chromatograph, featuring its own injection

system, so as to function without heated transfer lines and

without the need to occupy an injection port of the chro-

matograph. Other than the thermal desorption module, it

includes a refocusing trap based on freezing with liquid

nitrogen in order to enhance the chromatographic separation.

Its autonomous injection system provides the ability to inject

the analytes desorbed from the refocusing step in split

(injector purge) or splittless mode (no injector purge)

depending on the load of analytes to the system or the

detection limits required for the current analysis.

Desorption flow of He was set at 30 mL/min, while the

temperature was kept constant at 200 8C. Desorption and

refocusing duration was 20 min in order to maximize recov-

ery. The cryo trap capillary was a 22 cm part of a 0.53 mm

id, AT-Q, Q-PLOT column (Alltech Associates); it was

chosen in order to enhance trapping of ultra-VOCs and

consequently the chromatographic resolution. A 20 s heating

pulse has proved to be adequate for flush desorption of

trapped analytes in the GC column; short enough to prevent

extensive deterioration of the cryo trap column thereby

limiting artifacts in the analysis. Cryogen used was liquid

nitrogen.

A 60 m SPB-624 capillary column with 1.4 mm station-

ary phase and an internal diameter of 0.25 mm (Supelco)

was utilized for high-resolution chromatographic separation.

Column head pressure of helium purge gas was set to 25 psi.

GC program was selected as follows: 35 8C initial for 5 min,

ramp of 4 8C/min up to 180 8C, hold for 20 min. MSD mass

range was limited from 35 to 350 amu due to the expected

detection of VOCs, but with the benefit of 1.8 scans s�1.

3.5. Data processing

Chromatographic peaks were initially identified with the

help of Wiley 138 library, using similarity indexes higher

than 80%. In order to enhance peak identification, a sub-

stance database was constructed using ‘‘Easy-Id’’ tool of HP

Productivity Chemstation. A compound database for VOC

analysis has been used. The compound database consists of

the compounds recorded with reference samples and/or the

original ones. The record for each substance in the database

has several fields. Among them the retention time, most

abundant mass and three qualifier ions are the most impor-

tant. Quantitative results were generated with the application

of internal standard method [16], using the following equa-

tion:

Ci ¼Ai

AISTD� 1

RRFi�CISTD

where Ci = concentration of substance i in the vapor phase

expressed in nmol/L, Ai = peak area of substance i, in counts,

AISTD = peak area of internal standard, in counts calculated

for each run, RRFi = (Ai/AISTD) � (Ci/CISTD), relative

response factor of substance i, from standard mixture,

relative to chlorobenzene-d5; expresses the different ioniza-

tion of substance i compared to the internal standard,

CISTD = calculated concentration of internal standard in

the total air volume sampled in nmol/L.

The concentration calculation for the internal standard

has been done by reducing the known molar quantity


injected on the tube to the 5 L of air sampled. This con-

centration was calculated to be 10 nmol/L. This semi-quan-

titative full screening method was selected in order to offer

fast and reliable quantitative results in a variety of complex

substances without the need of constructing many calibra-

tion curves.

4. Results

Over 15 volatile substances were identified common in

all sampling experiments employed; over 60 substances

from the first two experiments. A variety of substances have

been identified, including almost all types of organic sub-

stances: hydrocarbons, aldehydes, ketones, alcohols, esters,

aromatics and sulfides. Among them dimethyl disulfide,

toluene, hexane, benzene 1,2,4-trimethyl, 2-propanone, 3-

pentanone, 2-pentanone and 2-methyl pentane showed the

highest concentrations. Table 1 presents the concentrations

(semi-quantitative analysis) of identified volatile substances

along with the mean retention time. A typical chromatogram

of VOCs identified during putrefaction is presented in Fig.

1a. Fig. 1b shows the reference sample taken at a close

distance. The mean ambient temperature measured during

sampling for cadaver one was 13.75 8C and the relative

humidity 69.32%. For cadaver 2 the mean temperature was

17.45 8C and humidity 51.37%, respectively. The ambient

temperature during the blank ‘‘body bag’’ SPME experiment

was 31.1 8C and humidity was 29.9%. The most prominent

VOCs identified migrating from the ‘‘body bag’’ were

phenol, 1-hexanol-2-ethyl and cyclohexanone.

5. Discussion

So far in the literature, there is not enough work to relate

VOCs with human decomposition. However, this has been

carried out in pathological conditions in living humans,

where similar experimental methodology has been widely

applied. Thus, in order to study VOCs during putrefaction

this type of available information was used in an attempt to

interpret the results of the present study.

It seems that there are three possible sources of VOCs in

this study. Atmospheric air trapped during packaging the

cadaver appears to be the first source. To examine this

possible contribution, a reference field blank sample was

taken outside the body bag and it did not show any measur-

able VOCs (Fig 1b). The next possible source is that VOCs

could migrate from the plastic ‘‘body bag’’ used to encase

the cadavers. This was tested by the SPME experiment

which was performed at a higher temperature than during

sampling. It appeared that phenol, 1-hexanol-2-ethyl and

cyclohexanone were found at very low concentrations.

Hexanol-2-ethyl has been detected in serum of hemodialysis

patients and in people suffering from liver disease [17]. It is

believed to be a metabolite of di-2-ethylhexyl phthalate

(DEHP), a widespread plasticizer often used in plastics

for medical use. Cyclohexanone found in the blank body

bag experiment was 50 times less than the amount found in

the air pumped in the experiments. It is then assumed that the

VOCs determined in the cadaver bags mainly evolved from

the dead bodies in putrefaction.

Dimethyl disulfide (DMDS) was found to be the sub-

stance with the highest concentration on both cadavers. It is

probably a byproduct of protein breakdown by bacteria.

Some studies in living humans suggest that the production

of many volatile sulfur compounds result from impaired

hepatic metabolism of sulfur-containing substances includ-

ing methionine [18].

DMDS, along with other mercaptanes, have been found

during headspace analysis in the urine of diabetic persons.

DMDS was supposed to result from enterobacterial degra-

dation of methionine in hepatic encephalopathy [19]. DMDS

was also identified in a study investigating sulfur compounds

from gingival crevicular sites in humans [20]. Generally

speaking, volatile sulfur compounds are commonly asso-

ciated with anaerobic decomposition of organic matter and

are well known for their characteristic strong disagreeable

and permeable odors in foods and drinks. Furthermore,

Ochiai et al. reported the identification of DMDS at low

ppbv in the expired air of humans using a three-stage

cryogenic trapping preconcentration system [21]. Tanger-

man et al. identified also DMDS in the breath of cirrhotic

patients [22]. Moreover, volatile sulfur compounds have

been suggested as one of the endogenous factors responsible

for hepatic coma [23]. Indicatively, the odor in liver failure

has been described as musty, rotten egg-like (corpse-like).

Although toluene was not predicted to be among the

most abundant substances, it was identified in both cases. It

has to be pointed out that the amount found was excep-

tionally high. The circumstances of death for both victims

are still unknown. One wonders if the large amount of

toluene is a result of poisoning. Inoue et al. identified

toluene in a headspace analysis of an adipoceratous body

mainly in adipose tissue, brain, skeletal muscle, liver and

kidney. Toluene is highly lipophilic and is thus found

abundantly in cadavers who previously were exposed to

toluene. Other volatiles also reported in their study and

also identified in the present work included dimethyl

disulfide, ethanol and 1-butanol [7]. Both alkyl benzenes

and trimethyl benzenes were found in low concentration

and this indicates that the deceased may have had some

contact with gasoline [24].

Acetone is produced by lipolysis during fat catabolism.

The breakdown of triglycerides in adipose tissue liberates

increased amounts of ketone bodies [25]. Acetone is formed

by decarboxylation of acetoacetate, which derives from

lipolysis or lipid peroxidation. High concentrations of acet-

one are found in expired air of living humans with uncon-

trolled diabetes mellitus [26]. In diabetic or starvation

ketoacidosis the breath has a characteristic odor that has

being described as the fruity aroma of old apples.


Table 1

Concentrations of VOCs identified on two decaying human males

Substance R̄ta (min) Concentration (nmol/L) Mean concentrationb (nmol/L)

Cadaver 1 Cadaver 2

1 Dimethyl disulfide 22.22 7.27 19.51 13.39

2 Toluene 23.10 6.85 13.37 10.11

3 Hexane 11.31 1.28 9.88 5.58

4 Benzene, 1,2,4-trimethyl 35.40 6.52 1.55 4.04

5 2-Propanone 7.97 2.04 5.64 3.84

6 3-Pentanone 19.84 0.95 6.22 3.59

7 2-Pentanone 19.41 1.70 4.90 3.30

8 Pentane, 2-methyl 9.31 0.17 5.04 2.61

9 Benzene, 1,3,5-trimethyl 34.90 2.54 – 2.54

10 Benzene, 1-ethyl, 3-methyl 34.02 2.71 2.06 2.39

11 Cyclopentane, methyl 12.99 – 2.11 2.11

12 Pentane, 3-methyl 10.12 0.07 3.64 1.86

13 Pentanal 19.58 – 1.78 1.78

14 Heptane, 2,4-dimethyl 23.19 – 1.68 1.68

15 Heptane 17.45 0.73 2.55 1.64

16 Benzene 1-ethyl, 2-methyl 35.07 1.83 1.40 1.62

17 Butanoic acid, ethyl ester 25.09 0.18 3.03 1.61

18 Hexanal 25.86 0.24 2.94 1.59

19 Octane 23.41 0.39 2.48 1.44

20 Benzene, 1,2,3-trimethyl 37.42 1.56 1.27 1.42

21 Ethanol 7.09 0.13 2.42 1.28

22 Propanoic acid, 2-methyl, ethyl ester 22.32 – 1.23 1.23

23 Furan, 2-ethyl 18.50 – 1.14 1.14

24 Disulfide, methyl ethyl 27.60 1.02 0.90 0.96

25 Carbon disulfide 8.30 0.47 1.22 0.85

26 1-Butanol 18.89 – 0.82 0.82

27 1-Heptene 17.13 0.36 1.20 0.78

28 Trisulfide, dimethyl 35.64 0.77 0.56 0.67

29 Benzene 16.55 0.98 0.32 0.65

30 2-Butanone 13.72 0.34 0.97 0.66

31 Acetic acid, propyl ester 20.30 0.29 0.96 0.63

32 m-Xylene 29.69 0.98 0.12 0.55

33 Cyclohexane 14.99 – 0.55 0.55

34 1-Hexene 11.16 0.36 0.67 0.52

35 Butane, 2,3-dimethyl 9.05 0.09 0.94 0.52

36 2-Hexanone 25.58 0.17 0.75 0.46

37 1-Hexanol 30.12 – 0.45 0.45

38 Acetic acid, ethyl ester 14.11 0.22 0.66 0.44

39 Benzene, propyl 33.71 0.42 0.33 0.38

40 p-Xylene 29.46 0.36 0.30 0.33

41 1-Pentanol 24.67 – 0.31 0.31

42 1-H-Indene, 2,3-dihydro 38.38 0.31 0.28 0.30

43 Pentane 6.13 0.02 0.52 0.27

44 2-Heptanone 30.78 – 0.26 0.26

45 Pentane, 2,4-dimethyl 12.75 0.16 0.33 0.25

46 Hexane, 2-methyl 15.61 0.24 – 0.24

47 Benzene, ethyl 28.56 0.28 0.09 0.19

48 Hexane, 3-methyl 15.94 0.20 0.17 0.19

49 Cyclohexanone 32.44 0.15 0.13 0.14

50 Benzene, 1,2-diethyl 38.25 – 0.12 0.12

51 Acetaldehyde 4.14 0.08 0.16 0.12

52 2-Pentene, 4-methyl 12.14 0.11 0.10 0.11

53 Butane, 2,2,3-trimethyl 12.53 0.05 0.14 0.10

54 Pentane, 2,3,4-trimethyl 20.75 0.06 0.13 0.10

55 Carbon oxide sulfide 2.80 0.04 0.15 0.10

56 2-Octene 24.02 0.06 0.12 0.09


Table 1 (Continued )

Substance R̄ta (min) Concentration (nmol/L) Mean concentrationb (nmol/L)

Cadaver 1 Cadaver 2

57 Alpha-pinene 31.77 0.08 0.09 0.09

58 Pentane, 2,3-dimethyl 15.28 – 0.08 0.08

59 Heptane, 3-methyl 21.34 – 0.08 0.08

60 Heptane, 2-methyl 21.13 – 0.07 0.07

61 Butanoic acid, 2-methyl ester 20.30 – 0.07 0.07

62 1-Pentanol, 2-methyl 20.81 – 0.07 0.07

63 Pentane, 2,3,3-trimethyl 21.04 0.07 0.06 0.07

64 Methane, chlorodifluoro 2.91 0.02 0.01 0.02

65 di-Limonene 36.75 – 0.06 0.06

66 Tetrachloromethane 15.73 0.08 0.04 0.06

67 3-Octene 25.90 – 0.05 0.05

68 Methane, chloro 3.38 0.08 0.02 0.05

69 2-Butylfuran 29.45 – 0.05 0.05

70 1-Propanol, 2-methyl 9.30 0.05 – 0.05

71 Disulfide, methyl propyl 33.34 0.04 – 0.04

72 Benzene, 1-ethyl, 4-methyl 32.33 0.04 – 0.04

73 Hexane, 2,3-dimethyl 20.90 – 0.04 0.04

74 Cyclohexane, 1,4-dimethyl 24.01 – 0.04 0.04

75 Octane, 2-methyl 26.90 – 0.03 0.03

76 Octane, 3,6-dimethyl 31.31 0.03 – 0.03

77 Butane, 2,2-dimethyl 7.41 – 0.02 0.02

78 Cyclohexane, 1-ethyl, 4-methyl 30.24 0.02 0.02 0.02

79 Isoprene 7.73 – 0.02 0.02

80 Cyclohexane, 1,2,4-trimethyl(1-alpha) 27.14 0.008 0.02 0.01

81 1-Pentene 5.80 – 0.01 0.01

82 1-Propene, 2-methyl 3.60 – 0.01 0.01

83 1-Propene 2.78 0.006 0.01 0.008

84 Cyclohexane, 1,3,5-trimethyl 26.02 0.008 – 0.008

85 Cyclohexane, 1,1,2,3-tetramethyl 33.08 – 0.007 0.007

86 1,3-Pentadiene 6.84 – 0.006 0.006

a Rt = retention time.b The mean concentration was taken in order to point out the most prominent VOCs.

Although the source of many VOCs evolved in humans is

not yet known, some of them have been correlated with

various pathological conditions and clearly related to meta-

bolic pathways in living humans: acetone from glucose

metabolism, alkanes from oxygen free radical-mediated

lipid peroxidation of fatty acids and isoprene from the

mevalonic acid pathway of cholesterol synthesis [27]. Some

VOCs that are produced in elevated concentrations in patho-

logical conditions in living humans are also present in

putrefaction, derived possibly from similar or different

metabolic pathways.

In the present study many alkanes and benzene deriva-

tives were found. Phillips et al. identified similar alkanes and

benzene derivatives in the breath of patients with lung cancer

[28]. As he states, part of the explanation may involve

increased oxygen free-radical activity in cancer cells. Oxy-

gen free radicals degrade cell membranes by lipid peroxida-

tion and convert these polyunsaturated fatty acids to volatile

alkanes. Alkanes are cleared from the body by excretion

through the lungs or by oxidation to alkyl alcohols by the

cytochrome P450 mixed-oxidase system.

Volatile fatty acids (i.e. propionic and butyric acid) are

primarily breakdown products of muscle and fat. Muscle

protein is degraded to component amino acids. Further

aerobic or anaerobic bacterial action leads to volatile fatty

acids. Further protein and fat decomposition yields phenolic

compounds and glycerols [13]. Although, these fatty acids

that are produced initially in the gut were not identified in the

present study, their esters were detected.

Many esters were also identified on both cadavers. These

included butanoic acid ethyl ester, propanoic acid 2-methyl

ethyl ester, acetic acid propyl ester, acetic acid ethyl ester

and butanoic acid 2-methyl ester. These are likely a result of

the saponification process. Saponification was more

advanced in the second cadaver. This was confirmed by

autopsy and volatile substance profile is in agreement (more

esters and volatiles in higher concentrations in cadaver 2).

Furthermore, CO2 concentration was greater in the sec-

ond cadaver by approximately 30%. In the first cadaver the

mean concentration of carbon dioxide during VOCs sam-

pling was 152 mmol/L whereas in cadaver 2 it was

196 mmol/L. However, measurements of ammonia concen-


Fig. 1. (a) A typical chromatogram of VOCs identified during putrefaction. The numbers in peaks are presented as indicated in Table 1; (b)

reference sample loaded with 1 mL internal standard taken at a close distance.

tration were not possible, due to limitations of the portable

gas sensor used (diffusion and sensitivity limit 0–100 ppm).

Ammonia, if present, in both cadavers was probably below

the detection limit of the sensor (<1 ppm). It should be

pointed out that the odor of ammonia is sharp and pungent.

Carbon dioxide and ammonia along with other gases (H2S,

CH4, SO2, H2) are present during putrefaction with resultant

distension of tissues. It should be emphasized that hydrogen

sulfide odor is like rotten eggs and sulfur dioxide odor is

pungent and irritating.

Tomita reported that dead bodies placed under water for

experimental purposes produced various alcohols such as

methanol, ethanol, 1-propanol, 1-butanol, occasionally 2-

propanol and others [29]. Furthermore, O’Neal et al. in a

review article reported a variety of volatiles found in animal

and human corpses under a variety of conditions: ethanol,

acetone, acetaldehyde, n-butanole, sec-butanole, n-propa-

nol, isopropanol, isoamyl alcohol, isobutanol, isopentanol,

ethyl ether, formaldehyde, phenyl ethanol and p-hydroxy-

phenylethanol. As he stated, the alcohols produced are

dependent on the bacteria present and the substrates avail-

able [30]. The first five volatile substances were also identi-

fied in the present study.

Other volatile substances reported in the literature and

considered significant during decomposition are indole and

skatole (fecal, nauseating odors), cadaverine (putrid, decay-

ing flesh odor) and putrescine (putrid, nauseating odor).

These were not detected in the present study, possibly due

to their low volatility [11]. Putrescine can be formed from

the decarboxylation of L-ornithine and cadaverine from

decarboxylation of L-lysine [31].

In this kind of experiments, environmental conditions are

critical and need to be considered. Temperature and humid-

ity not only affect the breakdown of proteins and carbohy-

drates, but may also modify the function of insects and

bacteria [13]. Huckenbeck found different bacterial species

producing different profiles of putrefactive volatiles. For

example C. aminovalericum produced acetone, ethanol and

1-butanol, whereas C. cadaveris produced acetone, ethanol,

1-butanol, isobutanol, n-butanol, isoamyl alcohol and n-

amyl alcohol. Temperature and humidity during putrefaction

will influence the profile of volatiles produced postmortem

[32]. However, volatile substances found in the present study

were identified under the given conditions (i.e. victims in the

sea, winter). Furthermore, volatile esters were likely formed

due to the presence of water.

6. Conclusion

The present study identifies volatile organic compounds

from two cadavers during putrefaction and raises a possible

role for them in forensic science. Over 80 VOCs were

identified and quantified. Among them dimethyl disulfide,


toluene, hexane, benzene 1,2,4-trimethyl, 2-propanone and

3-pentanone were found in higher concentrations. Although

this is one of the first systematic analyses of VOCs produced

during putrefaction from the entire body, it has to be inter-

preted in relation to the given conditions where the two men

were found in sea water. It is of note that these findings

represent measurements at a single time point in the decay

process. Increased esters found in the second cadaver may be

consistent with advanced saponification. Further qualitative

and quantitative study of volatile organic compounds pro-

duced under different known conditions may provide infor-

mation that can be used in the future to more accurately

estimate the postmortem interval for death cases under

unknown situations. SPME and other similar methods could

be used for further investigations of the putrefaction che-

mical profile.

References

[1] M. Fleisher, E. Simon, E. Rumpel, H. Ulmer, M. Harbeck, M.

Wandel, C. Fietzek, U. Weimar, H. Meixner, Detection of

volatile compounds correlated to human diseases through

breath analysis with chemical sensors, Sens. Actuat. B 4189

(2002) 1–5.

[2] Y.J. Lin, H.R. Guo, Y.H. Chang, M.T. Kao, H.H. Wang, R.I.

Hong, Application of the electronic nose for uremia diagnosis,

Sens. Actuat. B 76 (2001) 177–180.

[3] L.A. Wallace, E.D. Pellizzari, Personal air exposures and

breath concentrations of benzene and other volatile hydro-

carbons for smokers and non-smokers, Toxicol. Lett. 35 (1986)

113–116.

[4] M. Ogata, H. Michitsuji, Y. Fujiki, Estimating amounts of

toluene inhaled by workers with protective mask using biolo-

gical indicators of toluene, Toxicol. Lett. 108 (1999) 233–239.

[5] M. Abalos, X. Prieto, J.M. Bayona, Determination of volatile

alkyl sulfides in wastewater by headspace solid-phase micro-

extraction followed by gas chromatography-mass spectrome-

try, J. Chromatogr. A 963 (2002) 249–257.

[6] D.J. Tranthim-Fryer, R.C. Hansson, K.W. Norman, Headspace/

solid-phase microextraction/gas chromatography–mass spec-

trometry: a screening technique for the recovery and identi-

fication of volatile organic compounds (VOCs) in postmortem

blood and viscera samples, J. For. Sci. 46 (4) (2001) 934–946.

[7] H. Inoue, M. Iwasa, Y. Maeno, H. Koyama, Y. Sato, R. Matoba,

Detection of toluene in an adipoceratous body, For. Sci. Int. 78

(1996) 119–124.

[8] R.E. Zumwalt, R.O. Bost, I. Sunshine, Evaluation of ethanol

concentrations in decomposed bodies, J. For. Sci. 27 (1982)

549–554.

[9] P. Kulshrestha, D.K. Satpathy, Use of beetles in forensic

entomology, For. Sci. Int. 120 (2001) 15–17.

[10] C.P. Campobasso, F. Introna, The forensic entomologist in the

context of the forensic pathologist’s role, For. Sci. Int. 120

(2001) 132–139.

[11] A.A. Vass, S.A. Barshick, G. Sega, J. Caton, J.T. Skeen, J.C.

Love, J.A. Synstelien, Decomposition chemistry of human

remains: a new methodology for determining the postmorterm

interval, J. For. Sci. 47 (3) (2002) 542–553.

[12] W.C. Rodriquez, W.M. Bass, Decomposition of buried bodies

and methods that may aid in their location, J. For. Sci. 30 (3)

(1985) 836–852.

[13] A.A. Vass, W.M. Bass, J.D. Wolt, J.E. Foss, J.T. Ammons,

Time since death determinations of human cadavers using soil

solution, J. For. Sci. 37 (5) (1992) 1236–1253.

[14] E.A. Woolfenden, W.A. McClenny (Eds.), Compendium of

Methods for the Determination of Toxic Organic Compounds

in Ambient Air, 2nd ed., Compendium Method TO-17, Deter-

mination of Volatile Organic Compounds (VOCs) in Ambient

Air Using Active Sampling onto Sorbent Tubes, EPA, 1997,

pp. 17–19.

[15] H.J. Rafson, Odor and VOC Control Handbook, McGraw-Hill,

1998 pp. 4.24–4.26.

[16] W.A. McClenny, M.W. Holdren (Eds.), Compendium of

Methods for the Determination of Toxic Organic Com-

pounds in Ambient Air, 2nd ed., Compendium Method

TO-15, Determination of Volatile Organic Compounds

(VOCs) in Air Collected in Specially-prepared Canisters

and Analyzed by Gas Chromatography/Mass Spectrometry,

EPA, 1997.

[17] M.A. Faouzi, T. Dine, B. Gressier, K. Kambia, M. Luyckx, D.

Pagniez, C. Brunet, M. Kazin, A. Belabed, J.C. Cazin, Expo-

sure of hemodialysis patients to di-2-ethylhexyl phthalate, Int.

J. Pharmaceut. 180 (1999) 113–121.

[18] S.M. Studer, J.B. Orens, I. Rosas, J.A. Krishnan, K.A. Cope, S.

Yang, J.V. Conte, P.B. Becker, T.H. Risby, Patterns and sig-

nificance of exhaled-breath biomarkers in lung transplant

recipients with acute allograft rejection, J. Heart Lung Trans-

plant. 20 (11) (2001) 1158–1166.

[19] H.G. Wahl, A. Hoffmann, D. Luft, H.M. Liebich, Analysis of

volatile organic compounds in human urine by headspace gas

chromatography–mass spectrometry with a multipurpose sam-

pler, J. Chromatogr. A 847 (1999) 117–125.

[20] J.M. Coli, J. Tonzetich, Characterization of volatile sulfur

compounds production at individual gingival crevicular sites

in humans, J. Clin. Dent. 3 (4) (1992) 97–103.

[21] N. Ochiai, M. Takino, S. Daishima, D.B. Gardin, Analysis of

volatile sulfur compounds in breath by gas chromatography–

mass spectrometry using a three-stage cryogenic trapping

preconcentration system, J. Chromatogr. B 762 (2001)

67–75.

[22] A. Tangerman, M.T. Meuwese-Arends, J.H.M. van Tongeren,

A new sensitive assay for measuring volatile sulfur compounds

in human breath by Tenax trapping and gas chromatography

and its application in liver cirrhosis, Clinica Chim. Acta 130

(1983) 103–110.

[23] L. Zieve, W.M. Doizaki, F.J. Zieve, Synergism between mer-

captanes and ammonia or fatty acids in the production of coma:

a possible role for mercaptanes in the pathogenesis of hepatic

coma, J. Lab. Clin. Med. 83 (1974) 16–27.

[24] K.P. Kirkbride, S.M. Yap, S. Andrews, P.E. Pigou, G. Klass,

A.C. Dinan, F.L. Peddie, Microbial degradation of petroleum

hydrocarbons: implications for arson residue analysis, J. For.

Sci. 37 (6) (1992) 1585–1599.

[25] S.M. Kundu, J.A. Bruzek, R. Nair, A.M. Judilla, Breath

acetone analyzer: diagnostic tool to monitor dietary fat loss,

Clin. Chem. 39 (1) (1993) 87–92.

[26] J.K. Schubert, W.P.E. Muller, A. Benzing, K. Geiger, Applica-

tion of a new method for analysis of exhaled gas in critically ill

patients, Intensive Care Med. 24 (1998) 415–421.


[27] M. Phillips, J. Herrera, S. Krishnan, M. Zain, J. Greenberg,

R.N. Cataneo, Variation in volatile organic compounds in the

breath of normal humans, J. Chromatogr. B 729 (1999) 75–

88.

[28] M. Phillips, K. Gleeson, J.M.B. Hughes, J. Greenberg, R.N.

Cataneo, L. Bager, W.P. McVay, Volatile organic compounds in

breath as markers of lung cancer: a cross-sectional study,

Lancet 353 (1999) 1930–1933.

[29] K. Tomita, On putrefactions and floatations of dead

bodies under water, Hiroshima J. Med. Sci. 24 (1975)

117–152.

[30] C.L. O’ Neal, A. Poklis, Postmortem production of ethanol and

factors that influence interpretation: a critical review, Am. J.

For. Med. Pathol. 17 (1) (1996) 8–20.

[31] M. Cooke, N. Leeves, C. White, Time profile of putrescine,

cadaverine, indole and skatole in human saliva, Arch. Oral

Biol. 48 (2003) 323–327.

[32] W. Huckenbeck, Products of bacteria fermentation, in: W.

Bonte (Ed.), Congener Alcohols and their Medicolegal Sig-

nificance, Proceedings of the International Workshop on Coge-

ners of Alcoholic Beverages, Dusseldorf, 1987, pp. 27–36.

Documents

Volatiles Evolved From Decaying Human