Electronic nose and electronic mucosa as innovative instruments for real-time monitoring of food dryers

Accepted Manuscript Electronic nose and electronic mucosa as innovative instruments for real-time monitoring of food dryers Mahdi Ghasemi-Varnamkhasti , Mortaza Aghbashlo PII:

S0924-2244(14)00108-3

DOI:

10.1016/j.tifs.2014.05.004

Reference:

TIFS 1548

To appear in:

Trends in Food Science & Technology

Please cite this article as: Ghasemi-Varnamkhasti, M., Aghbashlo, M., Electronic nose and electronic mucosa as innovative instruments for real-time monitoring of food dryers, Trends in Food Science & Technology (2014), doi: 10.1016/j.tifs.2014.05.004. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Viewpoint

1 2

Title:

RI PT

3

Electronic nose and electronic mucosa as innovative

5

instruments for real-time monitoring of food dryers

6

SC

4

Authors:

8

Mahdi Ghasemi-Varnamkhasti 1*, Mortaza Aghbashlo 2

9

Affiliation:

11

1

Department of Mechanical Engineering of Biosystems, Shahrekord University, Shahrekord, Iran 2

Department of Agricultural Machinery Engineering, Faculty of Agricultural Engineering and

12

Technology, University of Tehran, Karaj, Iran

13

16 17 18

Corresponding author: Mahdi Ghasemi-Varnamkhasti

EP

15

*

Tel.: +98 3814424428

fax: +98 3814424428

E-mail address: [email protected]

AC C

14

TE D

10

M AN U

7

19 20

1

ACCEPTED MANUSCRIPT

21 Abstract

23

A crucial aspect of drying process is monitoring of product moisture content, nutritional

24

value, and sensorial characteristics. This viewpoint paper considers the challenges and the

25

opportunities in application of electronic nose and electronic mucosa as innovative

26

instruments in drying technology in order to improve the final food quality. Electronic

27

nose is reliable and easy to use tool in actual drying conditions. Nonetheless, the

28

electronic mucosa, imitates nasal chromatograph effect, can provide more useful

29

information leading to higher level of recognition than the existing electronic nose

30

systems. Perspectives and potential employment of such tools are discussed as well.

M AN U

SC

RI PT

22

31

Keywords: Food drying; Electronic mucosa; Electronic nose; Real-time monitoring

33

TE D

32

Background

35

Drying is a process widely used by the food industries to hinder the growth of

36

microorganisms, to guarantee the microbial stability, to make easy the handling of

37

product, and to assure the expected shelf-life of product (Aghbashlo, Kianmehr, &

38

Arabhosseini, 2008; Aghbashlo, Mobli, Rafiee, & Madadlou, 2013). Unfortunately, this

39

process might occasionally lead to low quality dried foods with unpleasant sensorial

40

properties worth mentioning due to unwanted physiochemical changes caused by

41

inappropriate drying conditions. Furthermore, the importance of food organoleptic

42

characteristics as first quality indicator for consumers encourage the food technologists to

43

highlight the better preserving and improving the foodstuffs original composition,

AC C

EP

34

2

ACCEPTED MANUSCRIPT

nutritional value, and functional properties during drying process. The aroma

45

volatilization and off-odors production are the major reasons for the food deterioration

46

during drying process, and, these reactions might alter the sensorial quality, nutritional

47

value, and functional properties, influencing negatively on the consumer preference.

48

Thus, the original aroma of dried product should be sufficiently kept to maintain its

49

functionality and marketability.

50

Although the sensorial properties of dried products such as aroma is the key element for

51

real-time process monitoring and consequently dryers controlling and monitoring, real-

52

time, fast, reliable, accurate, and cost-effective aroma profiling techniques are lacking.

53

Conventionally, aroma measurement of foodstuffs underlying different unit operations

54

was tremendously challenging and was generally confined to off-line, time-consuming,

55

and less efficient laboratory testing approaches. These techniques were usually based on

56

randomly collected samples, which limited their use for real-time monitoring and

57

accurate in-process closed-loop control. On the other hand, drying process is relatively

58

complex, dynamic, non-linear, uncertain, and unforeseeable operation because of

59

concurrent heat and mass transfer, fast moisture evaporation, case-hardening, and

60

intensive chemical and biochemical reactions (Aghbashlo, Mobli, Rafiee, & Madadlou,

61

2012). This complexity becomes even more in the case of food drying pronounced due to

62

product heterogeneity, anisotropicity, and non-uniformity (Nazghelichi, Aghbashlo,

63

Kianmehr, & Omid, 2011). Hence, empirical models based on off-line measurement

64

techniques cannot be precisely used for real-time controlling of drying process. As well,

65

more food industries move to continuous processes, in-process monitoring that provides

66

real-time organoleptic characteristics and feedback control that becomes even more

AC C

EP

TE D

M AN U

SC

RI PT

44

3

ACCEPTED MANUSCRIPT

necessary to produce high-nutritive and cost-effective dried foods. Moreover, tight food

68

process control is gained by reliable and precise controlling and monitoring of the process

69

factors influencing on the food sensorial properties. However, in drying process, these

70

factors are prone to natural inevitable changes, causing variability and difficult to foresee

71

the sensorial attributes of final product. It is also worth mentioning the quality of

72

foodstuffs should be examined during whole process and not only at the end-point. Thus,

73

food unit operations should be exactly monitored and controlled in whole processing

74

duration by use of innovative instruments to guarantee the quality of end products. These

75

are why developing an automatic real-time aroma tracking unit in food dryers is strictly

76

substantial to gain rigorous product sensorial attributes and reach a better understanding

77

of the process for careful optimization of the standard operating approaches.

78

On the other hand, according to Quality by Design (QbD) philosophy (Drennen, 2007),

79

quality cannot be assessed on a product but must be made and maintained into products

80

from the beginning of a process. According to Food QbD framework, sources of

81

variability in the drying process should be identified and hindered for improving the

82

quality and uniformity of dried products. This can be feasible by application of real-time

83

monitoring devices and techniques with feedback control through in-process

84

measurements (Fig. 1A). In the most of foodstuff drying processes, variability origins

85

affecting on final product quality and homogeneity can be specified, monitored, and even

86

controlled, according to Fig. 1B. Firstly, it is required to find out the relation between

87

drying process parameters and product quality characteristics and then the origins of

88

process variability must be monitored in timely procedure by process analyzers for

89

precise and efficient process control (Fig. 1A). Therefore, newly designed ‘‘smart’’ food

AC C

EP

TE D

M AN U

SC

RI PT

67

4

ACCEPTED MANUSCRIPT

dryers based on QbD framework might be more reliable and cost-effective than the

91

common ones.

92

Evidently, real-time measurement plays a key role in QbD scheme and according to Food

93

and Drug Administration (FDA, 2004), real-time process measurements can be classified

94

into three categories such as at-line, on-line, and in-line analyzers, as illustrated in Fig. 2.

95

In brief, at-line analyzers characterize the favorable properties by withdrawing the

96

product from process, isolating from ambient and analyzing it in close affinity to the

97

process stream. On-line analyzers specify the desired attributes by diverting the sample

98

from the process and returning to the process stream in most cases. In-line analyzers

99

(invasive or non-invasive) are rapid measuring tools or probes to consider the desired

100

characteristics without sample withdrawing by direct placing of them into a process

101

stream. Generally, implementation of real-time measurement units to drying equipments

102

should eventually result in control of the process with the least dried product testing and

103

in-depth process understanding assure the end product quality.

104

As mentioned earlier, a more useful way to monitor and control the drying process

105

consists of the real-time foodstuffs quality evaluation to get an actual insight about the

106

variables concerning both process and product. Hence, drying services should be

107

implemented with in-process automatic drying parameter manipulations to reduce the

108

undesirable changes in the product properties and process progress. In recent years,

109

electronic nose have been increasingly considered as an option for real-time

110

measurements of sensorial attributes of processed foods, as this instrument provides

111

reliable, non-invasive, and non-destructive measurements without sample withdrawing. It

112

can be also employed in a non-intrusive mode to build up qualitative and quantitative

AC C

EP

TE D

M AN U

SC

RI PT

90

5

ACCEPTED MANUSCRIPT

strategies for the in-process monitoring and controlling of a food drying process.

114

However, the advanced version of electronic nose namely electronic mucosa presents

115

numerous advantages over electronic nose due to its additional functionality originating

116

from its biological character. Stability, precision, reliability, rapid response, and

117

robustness of controller are the main specifications of any industrial dryer control system

118

(Jumah, Mujumdar, & Raghavan, 2006), which can be obtained by both instruments.

119

Instrumentation of food dryers with advanced control tools reduces batch failure,

120

improves reliability of process, minimizes labor cost, optimizes energy efficiency,

121

increases predictability of quality, and augments the finished product quality and

122

uniformity.

M AN U

SC

RI PT

113

123

Electronic nose and electronic mucosa and their potential applications

125

in drying technology

126

In most cases, food aroma contains complex mixtures of many volatile organic

127

compounds (VOC) with various sensorial and chemical attributes. Any change in the

128

relative concentration amounts of these compounds can often specify the odor of the

129

product. Nonetheless, not all the volatile compounds play an identical role in sample

130

aroma (Peris & Escuder-Gilabert, 2009; Shafiee, Minaei, Moghaddam, Ghasemi-

131

Varnamkhasti, & Barzagar, 2013). The influence of each compound is a function of its

132

concentration and its sensorial threshold (the minimum concentration perceivable by the

133

human nose). The ratio between these two items is known as the odor activity value

134

(OAV). When the OAV of a compound is more than 1, it will affect on food aroma

135

(Mistry, Reineccius, & Olson, 1997).

AC C

EP

TE D

124

6

ACCEPTED MANUSCRIPT

The change of food aroma during drying process is of great concern to the food

137

technologists who have been spent much funds in the monitoring and automating of their

138

plants for the foodstuffs drying. Nevertheless, variables which are generally considered at

139

drying facilities (moisture and temperature of the inlet and outlet air, temperature and

140

moisture content of products), do not present vision needed to specify the loss of

141

sensorial attributes such as aroma during processing. Undesired aroma loss, change,

142

distortion, or even destruction often occur and result in inappropriate quality of final

143

products. How to attain the desirable aroma and prevent unacceptable aroma is an

144

essential work for food engineers. Therefore, it could be particularly beneficial to

145

implement a real-time monitoring and controlling system in drying process for

146

measurement of the lost aroma in order to achieve a homogeneous and high quality

147

product. The aroma is caused by both enzymatic and non-enzymatic reactions as well as

148

degradations of macromolecules in the tissues of the food during the drying process.

149

Moreover, secondary metabolism of microorganisms is included in the creation of

150

volatile aroma components. Therefore, awareness on the odorant compounds at

151

processing stage is very crucial to successfully control of the quality of the food product.

152

To date, most research attempts on the aroma components of foods have been done with

153

complicated and luxurious tools such as gas chromatography (GC) and spectroscopy

154

methods. The big drawback of such techniques is that they cannot measure and quantify

155

the components in real-time and continuous mode. In spite of providing acceptable results

156

by such techniques, but analysis time is too long and usually require expert analysts and

157

technicians and therefore cannot be employed regularly. On the other hand, because of

AC C

EP

TE D

M AN U

SC

RI PT

136

7

ACCEPTED MANUSCRIPT

dynamic and complex nature of drying process, all of these instruments are inappropriate

159

for industrial applications due to extremely operator-dependency and poor repeatability.

160

In recent decade, the interests in employment of the most efficient methods to monitor the

161

food quality in drying technology have been emphasized (Mujumdar, 2007). Such

162

interests are concerning innovative technological advances, the increasing interest in the

163

enhancement of the quality of products, approve of more rules and standards for food,

164

and the establishment of R&D laboratories in the industry (Shirvani, Ghanbarian, &

165

Ghasemi-Varnamkhasti, 2014). All these circumstances have been led to make the

166

industry more quality conscious and so, it also results in the steady advance of quality

167

control in drying technology.

168

The electronic nose is an electronic system which tries to emulate the structure of the

169

biological nose (Fig. 3). The first stage for both biological and electronic noses is the

170

interaction among volatile compounds containing complex mixture with the suitable

171

receptors meaning a sensor array in electronic nose, and olfactory receptors in the case of

172

the biological nose. One odorant is recognized by multiple odorant receptors and one

173

odorant receptor is sensible to multiple odorants. The next stage is the saving of the

174

signals produced by the receptors in the brain or in a pattern recognition engine (learning

175

stage) and later the recognition of one of the odor saved (classification stage).

176

The output of the electronic nose is specified by the user and can be the characterizing of

177

an odorant, defect or attributes of the sample, concentration estimation for a certain

178

odorant, or the specifications of the odor as it might be perceived by a human, etc. In

179

electronic nose, each sensor in the array has overlapping sensitivities as shown in Fig. 4.

180

For instance, odorant No.1 may generate a low response in one sensor and higher

AC C

EP

TE D

M AN U

SC

RI PT

158

8

ACCEPTED MANUSCRIPT

responses in others, while odorant No.2 might generate high readings for sensors instead

182

of the one that “took” to odorant No.1. What should be taken into account is that the

183

pattern of response across the sensors is unique for diverse odorants. This distinguish

184

ability enables the system to characterize an unknown odor from the pattern of sensor

185

responses before learnt. The sensors in the array have distinctive response profiles to the

186

spectrum of odorants under study. The pattern of responses related to sensor array is

187

addressed as a fingerprint to recognize the odor.

188

Electronic nose is able to detect food aroma as rapid as in seconds, which makes real-

189

time monitoring of food aroma possible during drying for the first time. However, in

190

comparison to the off-line measurements, in-process measurement of organoleptic

191

characteristics is a concern in a food drying process. Electronic nose technology employs

192

an array of semi-selective gas sensors for identification of the volatile compounds present

193

in the headspace of a food sample (Lozano, Arroyo, Santos, Cabellos, & Horrillo, 2008;

194

Lozano, Santos, & Horrillo, 2008; Apetrei, Apetrei, Villanueva, de Saja, Gutierrez-

195

Rosales, & Rodriguez-Mendez, 2010; Ghasemi-Varnamkhasti, Mohtasebi, Siadat, 2010;

196

Ghasemi-Varnamkhasti, Mohtasebi, Siadat, Ahmadi, Razavi, & Dicko, 2011).

197

Advantages of electronic nose as innovative instrument consist of the fairly small amount

198

of sample preparation that is involved and the speed of analysis compared with traditional

199

tools like GC. Another merit of such instrument is that it is portable and can be used for

200

any drying systems without implementation issues.

201

However, a few reports have been documented in the literature on application of the

202

electronic nose for aroma measurement in food drying process. These research efforts

203

proved the use of electronic nose coupled with chemometric techniques as a reliable

AC C

EP

TE D

M AN U

SC

RI PT

181

9

ACCEPTED MANUSCRIPT

instrument to aroma monitoring of drying process. Future works goes to be focused on

205

using the electronic nose for monitoring of various food drying process and consequent

206

automating of drying plants employing advanced intelligent controlling approaches. Until

207

now, use of the electronic noses have kept limited just for off-line measurement (Yang et

208

al., 2009; Otero et al., 2003; Infante, Rubio, Contador, & Moreno, 2010; Lopez de

209

Lerma, Bellincontro, Mencarelli, Moreno, & Peinado, 2012; Laurienzo et al., 2013). It is

210

worth mentioning the manually gathered samples for off-line measurement at the end of

211

drying process are prone to changes in physical circumstances like humidity and

212

temperature, which will result in non-accurate aroma analysis.

213

More recently, few studies have been highlighted in the bibliography on zNose for on-

214

line monitoring of food in drying process (Li, Vijaya Raghavan, & Wang, 2010a; Li,

215

Vijaya Raghavan, & Wang, 2010b; Raghavan, Li, Wang, & Gariépy, 2010; Dev, Geetha,

216

Orsat, Gariépy, & Raghavan, 2011) and succeeding controlling based on volatile signals

217

(Li, Raghavan, Wang, & Gariepy, 2009). So far, there exists no report on application of

218

electronic nose for real-time monitoring of food dryers. However, real-time electronic

219

nose has an immense potential to application in drying technology for in-process

220

monitoring and controlling of food aroma apart from off-line electronic nose and

221

chemical methods. For instance, a schematic diagram of microwave drying system

222

together with zNose for aroma monitoring and controlling is depicted in Fig. 5 (Li et al.,

223

2009). It can be stated that the illustrated diagram can be successfully applied to the most

224

of drying equipments with minor changes. The components of the developed systems

225

were a volatiles detection unit, which in turn included a zNoseTM, a two-position

226

pneumatic valve, an air flow installation, a moisture condenser, and a flow meter.

AC C

EP

TE D

M AN U

SC

RI PT

204

10

ACCEPTED MANUSCRIPT

Nowadays, electronic nose generally faces with challenging drawbacks that confine their

228

application for real-time monitoring during drying process. Their sensitivity and detection

229

ability is greatly influenced by usual drift caused by temperature, humidity and

230

background noise, sensor variations, and sensor poisoning. Nevertheless, this innovative

231

instrument uses several sensors which are not very selective for special types of

232

compounds thus hindering any real recognition of any compound in the headspace of

233

food sample (Reid, O’Donnell, & Downey, 2006; Rock, Barsan, & Weimar, 2008). These

234

challenges, in addition to often wanting to measure very low concentrations (below ppm)

235

of the odor (Young, Buttner, Linnell, & Ramesham, 2003; Ryan et al., 2004), make the

236

employment of an electronic nose more complicated even with advanced auto samplers

237

and providing clean air. In close future when the main issues of the electronic nose have

238

been rectified, more real-time electronic noses will be designed and implemented in

239

drying equipments. For such applications, however, technical problems, as stated above,

240

have to be solved for real-time monitoring and controlling. Currently, many researchers

241

are working to overcome the problems limiting the capability of the electronic nose

242

(Wang, Tang, Chiu, Yang, & Kuo, 2011; Wali, 2012; Kaur, Kumar, Gulati, Ghanshyam,

243

Kapur, & Bhondekar, 2012; Bruins, Gerritsen, van de Sande, van Belkum, & Bos, 2013;

244

Zhanga, Tianb, Pengb, & Yin, 2014). The readers are referred to the review paper

245

complied by Gutiérrez and Horrillo (2014) in which the problems included on the

246

application of electronic nose for real-time cases and the solutions are discussed.

247

However, some instances of such problems are mentioned here:

248

1) Long sensors response time: to overcome this problem, another type of sensors could

249

be considered, i.e. acoustic wave sensors which show a faster response or even nanowires

AC C

EP

TE D

M AN U

SC

RI PT

227

11

ACCEPTED MANUSCRIPT

based gas sensors are faster, less time-consuming than typical thick or thin film gas

251

sensors. Employment of such sensor types for real-time application of electronic nose is

252

found in the literature (Shih, Lin, Lee, Chien, & Drake, 2012).

253

Moreover, it should be mentioned that the food dryers are generally available in batch

254

and continuous forms. Batch drying method is performed through loading the drying

255

chamber with material to be dried, developing the flow of drying media until adequately

256

drying of material, and then evacuating the dried product from the drying unit. On the

257

contrary, in continuous drying approach, the moist material as continuous flow is loaded

258

to the dryer and the dried material is also evacuated continuously. In the case of batch

259

drying mode, the drying duration is enough higher compared with the monitoring time-

260

frame of the electronic nose system. It is worth mentioning that the drying duration of

261

food is a function of many items such as material properties and drying conditions,

262

typically ranging from half of hour to several hours for most of moist foodstuffs. On the

263

other hand, continuous drying systems almost work at a quasi steady-state condition,

264

without regard to the short-term transient startup or shutdown time, and approximately

265

remain at a stable circumstance for a long time with the insignificant variations, until the

266

moment of abrupt changes in the drying medium and material conditions. These dryers

267

then quickly respond to a new drying condition and rapidly reach to the quasi steady-state

268

circumstance again. The quasi steady-state operating time of these dryers is very longer

269

than the monitoring time-frame of electronic nose sensors. Therefore, the monitoring time

270

window of electronic nose cannot be a very challenging and problematic subject to

271

hamper their successful application in food drying industry. Nonetheless, it is notable that

AC C

EP

TE D

M AN U

SC

RI PT

250

12

ACCEPTED MANUSCRIPT

the minimizing monitoring time window can result in more precise and extensive

273

application of electronic nose in drying technology.

274

2) High amount of training data (and hence long time and high costs of training):

275

currently many techniques are used to rectify this issue and chemometrics approaches

276

have shown promising trend. Principal Component Analysis (PCA) before training

277

reduces the redundancy and dimensionality of data, a remote training in servers

278

connecting the measurement computer with labview database tool kit. Moreover,

279

optimization techniques could be applied to select the temperatures of operation and the

280

composition of the array in order to obtain less overlapping in the selectivities of the

281

sensors. Poor selectivity is one of the weakest parameters of metal oxide semiconductor

282

(MOS) gas sensor as well as of most other types of sensors (Kolmakov, 2008) in such a

283

way to date, no a semiconductor with completely gas-specific response has been

284

fabricated. However, numerous approaches and techniques have been developed to

285

enhance the sensor selectivity (Zampolli, Elmi, Sturmann, Nicoletti, Dori, & Cardinali,

286

2005; Chen, Shen, & Zhou, 2008; Sysoev, Kiselev, Trouillet, & Bruns, 2013).

287

3) Lack of reliability in the long term (need of continuous recalibrations): To deal with

288

this problem, it can be stated that there are a lot of drift compensation techniques that

289

could be applied, even techniques for replacing sensors. Modeling sensors drift could also

290

be used for compensating its effect. Advanced approaches like on-line calibration have

291

been also reported in the literature (De Vito, Massera, Piga, Martinotto, & Di Francia,

292

2008; Geng, Yang, Wu, 2011; Zhang et al., 2011; Santonico et al., 2012). In recent

293

years, the techniques for reducing the time and cost of recalibration has been explored.

294

For instance, attenuation of the sensor failure (Pardo, Faglia, Sberveglieri, Corte, Masulli,

AC C

EP

TE D

M AN U

SC

RI PT

272

13

ACCEPTED MANUSCRIPT

& Riani, 2000; Fonollosa, Vergara, & Huerta, 2012; Fonollosa, Vergara, & Huerta,

296

2013), sensor poisoning (Padilla, Perera, Montoliu, Chaudry, Persaud, & Marco, 2007;

297

Reimann, Dausend, & Schutze, 2008, and sensor drift (Vergara, Vembu, Ayhan, Ryan,

298

Homer, & Huerta, 2012; Carlo, Falasconi, Sánchez, Scionti, Squillero, & Tonda, 2011;

299

Padilla, Perera, Montoliu, Chaudry, Persaud, & Marco, 2010; Ziyatdinov, Marco,

300

Chaudry, Persaud, Caminal, & Perera, 2010) has been aimed to extend the time between

301

recalibrations. More recently active control sampling has been reported in order to

302

minimize the number of experiments for recalibration (Rodriguez-Lujan, Fonollosa,

303

Vergara, Homer, & Huerta, 2014).

M AN U

SC

RI PT

295

A novel kind of electronic nose which tries to mimic nasal chromatograph effect with

305

more useful information content in such a way higher level of recognition compared with

306

the existing electronic nose systems could be developed (Taylor, Che Harun, Covington,

307

& Gardner, 2009; Liu, Ye, Yu, Hu, Cai, & Wang, 2009; Che Harun, Covington, &

308

Gardner, 2009; Liu et al., 2010; Che Harun, Covington, & Gardner, 2012). A typical

309

electronic mucosa as innovative instrument is shown in Fig. 6, in which three large arrays

310

of sensors with a two retentive columns have been combined (Che Harun, Taylor,

311

Covington & Gardner, 2009). In this instrument, all sensor arrays provide spatial

312

information (similar to common electronic nose, but with numerous sensors), but the

313

second and third arrays give temporal another profiles different from each other and the

314

first array.

315

As compared, the biological system employs many sensors with a various number of

316

binding proteins, coupled with a nasal chromatograph effect comparable to a GC column.

317

This further functionality enables the biological system to considerably act better than the

AC C

EP

TE D

304

14

ACCEPTED MANUSCRIPT

present electronic nose instruments (Beccherelli, Zampetti, Pantalei, Bernabei, &

319

Persaud, 2010). The biological system has the advantages such as higher selectivity,

320

sensitivity, and faster response than present electronic nose system based on chemical

321

sensor array (Dong, Du, Zhuang, Li, Liu, & Wang, 2013). In electronic mucosa, when an

322

odor signal passes through down a micro-channel, an absorbent coating provides a delay

323

akin to the work of the mucous layer in the nasal cavity. This absorbent material

324

selectively does a delay in the odor signal, consequently, the generation of an

325

odor/coating certain time delay. It is worth noting that the partitioning impact is to some

326

extent like a traditional gas (Gardner, Tan, Covington, & Pearce, 2007).

327

To date, no application of electronic mucosa in drying technology has been reported in

328

the literature, so this instrument could be very interesting to food technologists as a

329

promising trend in real-time monitoring of food drying in order to enhance the guarantee

330

and quality control of food products. The recent developments and achievements in the

331

field of sensory assessment and instrumental analysis as well as the interface between

332

humans (sensory science) and machines (instrumental analysis) show a bright future to

333

use the electronic mucosa in food dryers.

SC

M AN U

TE D

EP

334

RI PT

318

Conclusion and perspective

336

Aroma is one of the most important quality indicators of food products. This attribute, to

337

a great extent, specifies consumers’ preference and acceptance. However, in most food

338

processing operations, aroma can be lost, altered, distorted, or even destroyed. Retaining

339

desired aromas and limiting unpleasant aroma generation is a critical task and important

340

goal in food dryers. To fulfill this goal, food aroma should be properly and rapidly

AC C

335

15

ACCEPTED MANUSCRIPT

monitored and controlled during food processing in real-time mode. In other words, an

342

normal processing conditions of foodstuffs would contain in-line in-process organoleptic

343

attributes measurement and feedback control which can decrease human error, reduce

344

production cycle time, minimize analytical time and cost, augment material throughput

345

and provide improved sensory characteristics control. Electronic nose could be

346

considered to be used as real-time system for food quality monitoring in drying

347

technology. As well, the further advances in electronic nose and sensor technology along

348

with high capability of newer computerized system and new materials will show many

349

interesting and promising perspectives in monitoring, controlling and automating of food

350

dryers. One attractive vision for the future would be to have a fully automated platform

351

consisting of various kinds of sensors and instruments to monitor the substantial

352

information needed for the identification of quality of the raw material, process or

353

product (Ghasemi-Varnamkhasti et al., 2011; Ghasemi-Varnamkhasti & Forina, 2014).

354

Electronic nose would make up a very important part of such a multisensor system. The

355

artificial olfactory microsystem (electronic mucosa), as a novel technology, imitates the

356

fundamental structure of the mammalian olfactory system to provide spatial and temporal

357

chemosensory information. With regard to the advantages of electronic mucosa compared

358

with electronic nose, it is hoped that this innovative tool would be employed in food

359

drying in the close future. However, there is much research still to be carried out more

360

particularly with regard to sensors technology, data processing, analysis of results, and

361

validation studies. This may be realized in the drying technology in coming future.

AC C

EP

TE D

M AN U

SC

RI PT

341

362 363

16

ACCEPTED MANUSCRIPT

364

Acknowledgement

365

The authors would like to thank Dr. Lozano for constructive comments and kind

366

directions regarding this paper. The support of Shahrekord University is also appreciated.

RI PT

367 References

369

Aghbashlo, M., Kianmehr, M. H., & Arabhosseini, A. (2008). Energy and exergy

370

analyses of thin-layer drying of potato slices in a semi-industrial continuous band

371

dryer. Drying Technology, 26(12), 1501-1508.

SC

368

Aghbashlo, M., Mobli, H., Rafiee, S., & Madadlou, A. (2012). The use of artificial neural

373

network to predict exergetic performance of spray drying process: A preliminary

374

study. Computers and Electronics in Agriculture, 88, 32-43.

M AN U

372

Aghbashlo, M., Mobli, H., Rafiee, S., & Madadlou, A. (2013). A review on exergy

376

analysis of drying processes and systems. Renewable and Sustainable Energy

377

Reviews, 22, 1–22.

379

Anonymous. (2013). MEP Instruments Australia and New Zealand. . Available on http://mep.net.au/ processchemist, Dec 2013.

EP

378

TE D

375

Apetrei, C., Apetrei, I.M., Villanueva, S., de Saja, J.A., Gutierrez-Rosales, F., &

381

Rodriguez-Mendez, M.L., 2010. Combination of an e-nose, an e-tongue and an e-

382 383

AC C

380

eye for the characterisation of olive oils with different degree of bitterness. Analytica Chimica Acta, 663, 1, 91-97.

384

Beccherelli, R., Zampetti, E., Pantalei, S., Bernabei, M., & Persaud, K. C. (2010). Design

385

of a very large chemical sensor system for mimicking biological olfaction.

386

Sensors and Actuators B: Chemical, 146, 446–452.

17

ACCEPTED MANUSCRIPT

387

Bruins, M., Gerritsen, J. W., van de Sande, W., van Belkum, A., & Bos, A. (2013).

388

Enabling a transferable calibration model for metal-oxide type electronic noses.

389

Sensors and Actuators B: Chemical, 188, 1187– 1195. Carlo, S. D., Falasconi, M., Sánchez, E., Scionti, A., Squillero, G., Tonda, A. P. (2011).

391

Increasing pattern recognition accuracy for chemical sensing by evolutionary

392

based drift compensation. Pattern Recognition Letters, 32, 1594–1603.

394

Che Harun, F. K., Covington, J. A., & Gardner, J. W. (2009). Portable e-Mucosa System:

SC

393

RI PT

390

Mimicking the biological olfactory. Procedia Chemistry, 1, 991–994. Che Harun, F. K., Taylor, J. E.Covington, J. A., & Gardner, J. W. (2009). An electronic

396

nose employing dual-channel odor separation columns with large chemosensor

397

arrays for advanced odour discrimination. Sensors and Actuators B: Chemcial,

398

141, 134–140.

M AN U

395

Che Harun, F. K., Covington, J. A., & Gardner, J. W. (2012). Mimicking the biological

400

olfactory system: a portable electronic mucosa. IET Nanobiotechnology, 6(2), 45–

401

51.

TE D

399

Chen, P. C., Shen, G., Zhou, C. (2008). Chemical sensors and electronic noses based on

403

1D metal oxide nanostructures. IEEE Transactions on Nanotechnology, 7, 668–

404

682.

AC C

EP

402

405

De Vito, S., Massera, E., Piga, M., Martinotto, L., & Di Francia, G. (2008). On field

406

calibration of an electronic nose for benzene estimation in an urban pollution

407

monitoring scenario. Sensors and Actuators B: Chemical, 129, 750–757.

408

Dev, S. R. S., Geetha, P., Orsat, V., Gariépy, Y., & Raghavan, G. S. V. (2011). Effects of

409

microwave-assisted hot air drying and conventional hot air drying on the drying

18

ACCEPTED MANUSCRIPT

410

kinetics, color, rehydration, and volatiles of Moringa oleifera. Drying Technology,

411

29(12), 1452–1458. Dong, Q., Du, L., Zhuang, L., Li, R., Liu, Q., & Wang, P. (2013). A novel bioelectronic

413

nose based on brain–machine interface using implanted electrode recording in

414

vivo in olfactory bulb. Biosensors and Bioelectronics, 49, 263–269.

417 418

Pharmaceutical Innovation, 2(3), 65-66.

SC

416

Drennen, J. K. (2007). Quality by Design—What Does it Really Mean?. Journal of

Fonollosa, J., Vergara, A., & Huerta, R. (2012). Sensor failure mitigation based on multiple kernels. Sensors, IEEE, 1–4.

M AN U

415

RI PT

412

419

Fonollosa, J., Vergara, A., & Huerta, R. (2013). Algorithmic mitigation of sensor failure:

420

is sensor replacement really necessary? Sensors and Actuators B: Chemical, 183,

421

211–221.

Food and Drug Administration (FDA), Guidance: PAT – A Framework for Innovative

423

Pharmaceutical Development, Manufacturing and Quality Assurance, US

424

Department of Health and Human Services, 2004.

TE D

422

Gardner, J. W., Tan, S. L., Covington, J. A., & Pearce, T. C. (2007). Enhanced

426

discrimination complex odours based upon spatio-temporal signals from micro-

427

mucosa. The 14th International Conference on Solid-State Sensors, Actuators and

AC C

428

EP

425

Microsystems, Lyon, France, June 10–14, 2007.

429

Ghasemi-Varnamkhasti, M., Mohtasebi S. S., & Siadat, M. (2010) Biomimetic-based

430

odor and taste sensing systems to food quality and safety characterization: An

431

overview on basic principles and recent achievements. Journal of Food

432

Engineering, 100 (3), 377–387.

19

ACCEPTED MANUSCRIPT

Ghasemi-Varnamkhasti, M., Mohtasebi, S. S., Rodriguez- Mendez, M. L., Siadat, M.,

434

Ahmadi, H., & Razavi, S. H. (2011). Electronic and bioelectronic tongues, two

435

promising analytical tools for the quality evaluation of non alcoholic beer. Trends

436

in Food Science & Technology, 22, 245–248.

RI PT

433

Ghasemi-Varnamkhasti, M., Mohtasebi, S. S., Siadat, M., Ahmadi, H., Razavi, S. H., &

438

Dicko, A. (2011). Aging fingerptint characterization of beer using electronic nose.

439

Sensors and Actuators B: Chemical, 159, 51–59.

SC

437

Ghasemi-Varnamkhasti, M., Forina, M., (2014). NIR spectroscopy coupled with

441

multivariate computational tools for qualitative characterization of the aging of

442

beer. Computers and Electronics in Agriculture, 100, 34–40.

M AN U

440

Geng, Z., Yang, F., & Wu, N. (2011). Optimum design of sensor arrays via simulation-

444

based multivariate calibration. Sensors and Actuators B: Chemical, 156, 854–862.

445

Gutiérrez, J., & Horrillo, M. C. (2014). Advances in artificial olfaction: Sensors and

446

TE D

443

applications. Talanta, 124, 95–105.

Infante, R., Rubio, P., Contador, L., & Moreno, V. (2010). Effect of drying process on

448

lemon verbena (Lippia citrodora Kunth) aroma and infusion sensory quality.

449

International Journal of Food Science & Technology, 45(1), 75–80.

EP

447

Jumah, R. Y., Mujumdar, A. S. (1995). Control of industrial dryers. In A. S. Mujumdar

451

(Ed.), Handbook of Industrial Drying (pp.1343–1368). New York: Marcel

452

AC C

450

Dekker.

453

Kaur, R., Kumar, R., Gulati, A., Ghanshyam, C., Kapur, P., & Bhondekar, A. P. (2012).

454

Enhancing electronic nose performance: A novel feature selection approach using

455

dynamic social impact theory and moving window time slicing for classification

20

ACCEPTED MANUSCRIPT

456

of Kangra orthodox black tea (Camellia sinensis (L.) O. Kuntze). Sensors and

457

Actuators B: Chemcial, 166, 309–319. Kolmakov, A. (2008). Recent trends in the fabrication, functionalisation and

459

characterization of metal oxide nanowire gas sensors. International Journal of

460

Nanotechnology, 5, 450–474.

RI PT

458

Laurienzo, P., Cammarota, G., Di Stasio, M., Gentile, G., Laurino, C., & Volpe, M. G.

462

(2013). Microstructure and olfactory quality of apples de-hydrated by innovative

463

technologies. Journal of Food Engineering, 116(3), 689–694.

465

Li, Z., Vijaya Raghavan, G. S., & Wang, N. (2010a). Carrot volatiles monitoring and

M AN U

464

SC

461

control in microwave drying. LWT–Food Science and Technology, 43, 291–297. Li, Z., Vijaya Raghavan, G. S., & Wang, N. (2010b). Apple volatiles monitoring and

467

control in microwave drying. LWT–Food Science and Technology, 43(4), 684–

468

689.

TE D

466

Li, Z., Raghavan, G. S., Wang, N., & Gariepy, Y. (2009). Real-time, volatile-detection-

470

assisted control for microwave drying. Computers and Electronics in Agriculture,

471

69(2), 177–184.

EP

469

Liew, C. V., Wang, L. K., & Wan Sia Heng, P. (2010). Development of a visiometric

473

process analyzer for real-time monitoring of bottom spray fluid-bed coating.

474

AC C

472

Journal of Pharmaceutical Sciences, 99(1), 346–356.

475

Liu, Q., Ye, W., Yu, H., Hu, N., Cai, H., & Wang, P. (2009). Olfactory mucosa tissue

476

based biosensor for bioelectronic nose. Olfaction and Electronic Nose,

477

Proceedings of the 13th International Symposium 15–17 April, Brescia, Italy.

21

ACCEPTED MANUSCRIPT

478

Liu, Q., Ye, W., Yu, H., Hu, N., Du, L., Wang, P., & Yang, M. (2010). Olfactory mucosa

479

tissue-based biosensor: A bioelectronic nose with receptor cells in intact olfactory

480

epithelium. Sensors and Actuators B: Chemical, 158, 25–31. Lopez de Lerma, N., Bellincontro, A., Mencarelli, F., Moreno, J., & Peinado, R. A.

482

(2012). Use of electronic nose, validated by GC–MS, to establish the optimum

483

off-vine dehydration time of wine grapes. Food Chemistry, 130(2), 447–452.

485

Lozano, J., Arroyo, T., Santos, J.P., Cabellos, J.M., & Horrillo, M.C. (2008). Electronic

SC

484

RI PT

481

nose for wine ageing detection. Sensors and Actuators B, 133, 180-186. Lozano, J., Santos, J.P., Horrillo, M.C. (2008). Enrichment sampling methods for wine

487

discrimination with gas sensors. Journal of Food Composition and Analysis, 21,

488

716-723.

490 491 492

Mistry, B. S., Reineccius, T., Olson L. K., in: R. Marsili (Editor). (1997).Techniques for Analyzing Food Aroma, Marcel Dekker Inc, New York, USA, p. 265.

TE D

489

M AN U

486

Mujumdar, A. S. (2007). An overview of innovation in industrial drying: current status and R&D needs. Transportation in Porous Media, 66, 3–18. Nazghelichi, T., Aghbashlo, M., Kianmehr, M. H., & Omid, M. (2011). Prediction of

494

energy and exergy of carrot cubes in a fluidized bed dryer by artificial neural

495

networks. Drying Technology, 29(3), 295–307.

AC C

EP

493

496

Otero, L., Horrillo, M. C., Garcia, M., Sayago, I., Alexandre, M., Fernandez, M., Ares,

497

L., & Gutierrez, J. (2003). Detection of Iberian ham aroma by a semiconductor

498

multisensorial system. Meat Science, 65, 1175–1185.

22

ACCEPTED MANUSCRIPT

499

Padilla, M., Perera, A., Montoliu, I., Chaudry, A., Persaud, K., & Marco, S. (2010). Drift

500

compensation of gas sensor array data by orthogonal signal correction,

501

Chemometric and Intelligent Laboratory Systems, 100(1), 28–35. Padilla, M., Perera, A., Montoliu, I., Chaudry, A., Persaud, K., & Marco, S. (2007).

503

Poisoning fault diagnosis in chemical gas sensor arrays using multivariate statistical

504

signal processing and structured residuals generation, Intelligent Signal Processing,

505

WISP. IEEE International Symposium on, 2007, pp. 1–6.

SC

RI PT

502

Pardo, M., Faglia, G., Sberveglieri, G., Corte, M., Masulli, F., & Riani, M. (2000).

507

Monitoring reliability of sensors in an array by neural networks, Sensors and

508

Actuators, B: Chemical, 67, 128–133.

509 510

M AN U

506

Peris, M., & Escuder-Gilabert, L. (2009). A 21st century technique for food control: Electronic noses. Analytica Chimica Acta, 638, 1–15.

Raghavan, G. S. V., Li, Z., Wang, N., & Gariépy, Y. (2010). Control of microwave

512

drying process through aroma monitoring. Drying Technology, 28(5), 591-599.

513

Reid, L., O’Donnell, C., & Downey, G. (2006). Recent technological advances for the

514

determination of food authenticity. Trends in Food Science & Technology, 17, 344–

515

353.

517 518 519

EP

Reimann, P., Dausend, A., & Schutze, A. (2008). A self-monitoring and self-diagnosis

AC C

516

TE D

511

strategy for semiconductor gas sensor systems. Sensors, IEEE, 192–195.

Rock, F., Barsan, N., & Weimar, U. (2008). Electronic Nose: Current Status and Future Trends. Chemical Reviews, 108, 705–725.

23

ACCEPTED MANUSCRIPT

Rodriguez-Lujan, I., Fonollosa, J., Vergara, A., Homer, M., & Huerta, R. (2014). On the

521

calibration of sensor arrays for pattern recognition using the minimal number of

522

experiments. Chemometrics and Intelligent Laboratory Systems, 130, 123–134.

523

Ryan M. A., Homer M. L., Buehler M. G., Manatt, K. S. Lau, B. Karmon D., & Jackson,

524

S. (2004). Monitoring space shuttle air quality using the jet propulsion laboratory

525

electronic nose. IEEE Sensors Journal, 4(3), 337–347.

RI PT

520

Santonico, M., Pennazza, G., Capuano, R., Falconi, C., Vink, T. J., & Knobel, H. H., et

527

al., (2012). Electronic noses calibration procedure in the context of a multicentre

528

medical study. Sensors and Actuators B: Chemical, 173, 555–561.

M AN U

SC

526

529

Shafiee, S., Minaei, S., Moghaddam, N., Ghasemi-Varnamkhasti. M., & Barzegar, M.

530

(2013). Potential application of machine vision to honey characterization. Trends in

531

Food Science and Technology, 30, 174–177.

Shih, C., Lin, Y., Lee, K., Chien, P., & Drake, P. (2012). Real-time electronic nose based

533

pathogen detection for respiratory intensive care patients. Sensors and Actuators B:

534

Chemical, 148, 153–157.

TE D

532

Shirvani, M., Ghanbarian, D., & Ghasemi-Varnamkhasti, M. (2014). Measurement and

536

evaluation of the apparent modulus of elasticity of apple based on Hooke’s,

537

Hertz’s

and

AC C

538

EP

535

Boussinesq’s

theories.

Measurement,

doi:

10.1016/j.measurement.2014.04.014. In press.

539

Sysoev, V. V., Kiselev, I., Trouillet, V., & Bruns, M. (2013). Enhancing the gas

540

selectivity of single-crystal SnO2: Pt thin-film chemiresistor microarray by SiO2

541

membrane coating. Sensors and Actuators B: Chemical, 185, 59–69.

24

ACCEPTED MANUSCRIPT

Taylor, J. E., Che Harun, F. K., Covington, J. A., & Gardner, J. W. (2009). Applying

543

convolution-based processing methods to a dual-channel, large array artificial

544

olfactory mucosa. Olfaction and Electronic Nose, Proceedings of the 13th

545

International Symposium 15–17 April, Brescia, Italy.

RI PT

542

Wali, R. P. (2012). An electronic nose to differentiate aromatic flowers using a real-time

547

information-rich piezoelectric resonance measurement. Procedia Chemistry, 6,

548

194–202.

SC

546

Wang, L. C., Tang, K. T., Chiu, S. W., Yang, S. R., & Kuo, C. T. (2011). A bio-inspired

550

two-layer multiple-walled carbon nanotube–polymer composite sensor array and a

551

bio-inspired fast-adaptive readout circuit for a portable electronic nose.

552

Biosensors and Bioelectronics, 26, 4301– 4307.

M AN U

549

Vergara, A., Vembu, S., Ayhan, T., Ryan, M. A., Homer, M. L., Huerta, R. (2012).

554

Chemical gas sensor drift compensation using classifier ensembles. Sensors and

555

Actuators B: Chemical, 166, 320–329.

TE D

553

Yang, Z., Dong, F., Shimizu, K., Kinoshita, T., Kanamori, M., Morita, A., & Watanabe,

557

N. (2009). Identification of coumarin-enriched Japanese green teas and their

558

particular flavor using electronic nose. Journal of Food Engineering, 92, 312–316.

560

Young, R. C., Buttner, W. J., Linnell, B. R., & Ramesham, R. (2003). Electronic nose for

AC C

559

EP

556

space program applications. Sensors and Actuators B: Chemical, 93, 7–16.

561

Zampolli, S., Elmi, I., Sturmann, J., Nicoletti, S., Dori, L., & Cardinali, G. C. (2005).

562

Selectivity enhancement of metal oxide gas sensors using a micromachined gas

563

chromatographic column. Sensors and Actuators B: Chemical, 105, 400–406.

25

ACCEPTED MANUSCRIPT

Zhang, L., Tian, F., Kadri, C., Xiao, B., Li, H., Pan, L., & Zhou, H. (2011). On-line

565

sensor calibration transfer among electronic nose instruments for monitoring

566

volatile organic chemicals in indoor air quality. Sensors and Actuators B: Chemical,

567

160, 899–909.

RI PT

564

Zhang, L., Tianb, F., Pengb, X., & Yin, X. (2014). A rapid discreteness correction

569

scheme for reproducibility enhancement among a batch of MOS gas sensors.

570

Sensors and Actuators A: Physical, 205, 170–176.

SC

568

Ziyatdinov, A., Marco, S., Chaudry, A., Persaud, K., Caminal, P., & Perera, A. (2010).

572

Drift compensation of gas sensor array data by common principal component

573

analysis. Sensors and Actuators B: Chemical, 146, 460–465.

M AN U

571

574 575

579 580 581 582 583 584

EP

578

AC C

577

TE D

576

585 586

26

ACCEPTED MANUSCRIPT

587 588 Figure Captions

590

Fig. 1. Role of real-time monitoring in food quality by Design framework (with kind permission from

591

Liew, Wang, & Wan Sia Heng, 2010).

592

Fig. 2. A schematic representation of the difference among in-line, on-line, and at-line product

593

measurements during manufacturing (Anon. MEP Instruments Pty Ltd, NSW, Australia, 2013).

594

Fig. 3. A comparison of the biological and artificial olfactory systems.

595

Fig. 4. Sensitivities of gas sensors in an Electronic nose.

596

Fig. 5. Schematic diagram of the microwave drying system along with zNoseTM for aroma monitoring and

597

controlling (with kind permission from Li et al., 2009).

598

Fig. 6. Dual-column concept with large sensor arrays designed by Che Harun et al. (2009), (with kind

599

permission from Che Harun et al. 2009).

SC

M AN U

AC C

EP

TE D

600

RI PT

589

27

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

Fig. 1. Role of real-time monitoring in food quality by Design framework (with kind permission from Liew, Wang, & Wan Sia Heng, 2010).

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 2. A schematic representation of the difference among in-line, on-line, and at-line product measurements during

AC C

EP

TE D

manufacturing (Anon. MEP Instruments Pty Ltd, NSW, Australia, 2013).

Brain

Acquisition

Data Processing

Neurons

Human

SC

Smell

RI PT

ACCEPTED MANUSCRIPT

M AN U

Odor

Results

Neural Networks

Recognized as coffee

Processed Signal

AC C

Raw signal

EP

TE D

Comparison of odor by the neural system

Fig. 3. A comparison of the biological and artificial olfactory systems.

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

Fig. 4. Sensitivities of gas sensors in an Electronic nose

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 5. Schematic diagram of the microwave drying system along with zNoseTM for aroma monitoring and

AC C

EP

TE D

controlling (with kind permission from Li et al., 2009).

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Fig. 6. Dual-column concept with large sensor arrays designed by Che Harun et al. (2009), (with kind

AC C

EP

TE D

permission from Che Harun et al. 2009).

ACCEPTED MANUSCRIPT

* Electronic nose and electronic mucosa are reliable instruments; * Monitoring the changes in drying technology can provide suitable food quality;

RI PT

* Gas sensors are included in e-nose;

AC C

EP

TE D

M AN U

SC

* e-nose and e-mucosa coupled with NIR, e-tongue, and machine vision is a novel method.