Competing, complementary and co-existing paradigms in techno-scientific literature: A case study of Nanotechnology for engineering

Abstract

Nanotechnology is a research field that has potential to drive the progress of mankind for the next few decades. Its application is found in every discipline, ranging from material science to space communication. Owing to its potential for ubiquity, and capability of replacing many general purpose technologies, co-existence of several paradigms are expected in nanotechnology. Flow Vergence (FV) gradient has been recently introduced as a metric to mine the network of scientific literature for detecting the paradigm shifts. In this paper, we have performed citation network analysis of scientific publications in nanotechnology from research area ‘engineering’ for identification of paradigms related to the same. Flow vergence gradient revealed 18 subnetworks that deal with 25 likely pivots of paradigm shifts. Major paradigm shifts can be found in the field of targeted drug delivery. Nanonetworks, a crossover of IT, BT and nanotechnology is the another interesting paradigm shift identified. An extended subnetwork analysis has been conducted to identify the competing or complementary nature of the emerging paradigms in the subnetworks. A framework for this has also been introduced. This analysis revealed that most of the paradigms in the targeted delivery are competing paradigms. Complementary paradigms are also identified in nano electronics and targeted drug delivery. Policy implications from this identification for various target groups are also discussed.

Introduction

Nanotechnology is the generic name given to a set of technologies that deals with and manipulates materials with size in the nanometer scale (atoms and some molecules are found in the range of ‘nano’ size). It is a field with enormous potential that would revolutionize almost all the production and service industries. Like biotechnology and many other decisive technologies, the origin as well as early practical applications of nanotechnology can be traced back to prehistoric times. Modern day journey of nanotechnology as a field of research began in 1959, with Richard Feynman’s (Nobel laurete-1965) lecture titled as “There’s plenty of room at the bottom”, at the annual meeting of American Physical society at Caltech. Enormous opportunities likely to be brought up by miniaturization and challenges associated with it, as envisioned by Feyman, made the content of the talk. The transcript of the same has been published in 1960 as Feynman (1960). Over last 57 years, the field has evolved immensely. Many developed countries like US, Japan, Canada, China, Australia, UK and EU etc., have been formulating and amending separate nanotechnology policy as a part of their science-technology-innovation policies since 2000. Despite being a field which hoards as much challenges as the opportunities it offers, it is anticipated as one of the drivers of the 6th kondratieff cycle Pantin (2012), along with biotechnology.

Thomas Kuhn’s model of science dynamics based on paradigms and shifts in paradigms given in ‘structure of scientific revolutions’ Kuhn (1962) is highly relevant in this fast growing world. In his post script Kuhn (1969) to Kuhn (1962), Kuhn envisioned the chance for existence of multiple paradigms when there is an inevitable increase in pace and intensity of research activities (both scientific and technological research), which is a reality today. Dosi introduced the notion of ‘technological paradigms’ after observing the parallelism between ‘scientific paradigms’ and technological paradigms Dosi (1982). However, inherent unpredictability in innovation makes R&D investments a risky affair when it comes to technological research. Therefore understanding the threats like unexpected disruption of the existing technologies and opportunities like the rise of new technologies or sometimes the emergence of new technological paradigms is very vital for formulation of necessary policies. Érdi et al. (2013) addressed the problem of prediction of emerging technologies based on patent citation networks. Lathabai et al. (2015) and Prabhakaran et al. (2015) used scientific publications citation networks to detect the paradigm shifts and emerging fields from scientific literature. Prediction of paradigm shifts from scientific literature, which is of much greater gravity and difficulty than detection is addressed recently in Prabhakaran et al. (2018). Even though such predictions can forewarn researchers, investors and policymakers about scientific or technological paradigm shifts, a framework that could ease the process of identification of the nature of co-existence of several paradigms is imperative. This is discussed next.

Owing to its wide range of applicability and its ability to revolutionize many existing technologies including the general purpose technologies (GPTs), many paradigms are likely to co-exist with in nanotechnology. Identification of the paradigm shifts that led to the invasion of nanotechnology into many general purpose applications are investigation worthy. In some critical applications that have high impact over the life of people, like biomedical applications, several co-existing paradigms may compete with each other. Analysis of scientific literature is one of the best ways to identify these paradigm shifts. Therefore, in this work, our objective is to identify paradigm shifts (including the emerging ones) and the nature of the identified paradigms that co-exist in nanotechnology in the research area ‘engineering’. This is addressed through the analysis of scholarly publications from 1989 to 2015, by means of a recently identified phenomenon in scientific literature- ‘Flow Vergence effect’. In Lathabai et al. (2015), the metric ‘FV gradient’ was introduced to detect the flow vergence effect. However, it was employed only for path analysis of citation network of ‘Biotechnology for Engineering’. Here, the usage of this metric is extended to the whole network of ‘Nanotechnology’ for the research area ‘engineering’, to identify all the likely pivot papers of these paradigm shifts. Subnetworks that contain pivots were extracted and there were 25 pivots in 18 subnetworks. Thereafter, all the papers in these subnetworks are extended to their immediate neighbourhood, to visualize and investigate the clusters formed by these subnetworks. Based on the analysis of these clusters, nature of the co-existing paradigms (whether they are competing or complementary or merely co-existing ?) are identified based on the ‘dichromatic framework’ we have introduced.

FV gradient’s ability to detect pivot paper of paradigm shifts is demonstrated in its introductory paper and in this paper too. Content analysis of such papers provide indications to the nature of paradigm shifts. Thus in that sense the detected pivots pinpoint ‘likely paradigm shifts’. Analyst’s own expertise in respective fields and/or opinions from experts in the field is required to draw firm conclusions about the paradigm shifts. Insights from citation network analysis can be used as a base for conducting expert opinion or other judgmental studies or in combination with other data driven methods. In this context, our contribution towards citation network literature in the form of FV gradient analysis and dichromatic framework may unlock the hidden potential of citation network analysis for policy making applications.

Organisation of the article

Rest of this article is organised as follows. “Related literature” section is related literature on developments in citation network analysis, which form the integral part of the methodology adopted in this study. Also a review of attempts to study progress of nanotechnology using citation analysis and network approach is included in the same. “Methodology” section deals with the methodology and discusses how co-existing paradigm shifts can be identified in the citation network. “Results and discussions” section comprises of report and analysis of the results. This work is concluded in “Findings and policy implications” section after laying a roadmap for further research.

Related literature

Since industrial revolution, each era of technological progress were driven by one or a few general purpose technologies (GPTs) (Bresnahan and Trajtenberg 1995). In other words, each K-cycle or techno-economic paradigms have atleast one GPT as the key driver. For instance, steam engine was the main driver of first long wave of the K-cycle and information technology (IT) was the main driver of the fifth K-cycle. Works such as Helpman (1998) and Helpman and Trajtenberg (1998a, b) that extended the above mentioned seminal paper maintained that in most of the K-cycles, single GPT usually dominate the economy. This is contradictory to the real world experience as multiple GPT co-existence was witnessed (Carlaw and Lipsey 2011). The possibility of some of them to compete with others and also the possibility of some of them to complement others were also discussed in that work. An interesting exploration of the ways to identify the source and nature of a new GPTs in the not-too-distant future capable of supporting the current K-cycle as well as the next K-cycle (Coccia 2017) also favours the above argument. A significant extract from Dosi (1982) about the definition of a technological paradigm is “It would perhaps be better to talk of cluster of technologies, e.g. nuclear technologies, semiconductor technologies, organic chemistry technologies, etc”. Eventhough it is considered as a driver of the sixth K-cycle along with Biotechnology, Nanotechnology is not just a GPT. It can be viewed as a cluster of several GPTs and some enabling technologies or rather a cluster of such clusters. Therefore, it qualifies as a technological paradigm as defined by Dosi (1982) or a set of several technological paradigms. For instance, it could revolutionize semiconductor technologies and organic chemistry technologies causing the emergence of new technological paradigms such as nano semiconductor technologies and nano organic chemistry technologies. Thus, the invasion of nanotechnology is witnessing the phenomenon of co-existence of many technological paradigms which may be competing, complementary or merely co-existing in nature. As stated in the introduction, identification of such co-existing paradigms and their nature is of tremendous importance for many applications which include policy making. Mining of scientific literature and patent literature might offer a means for achieving the above said objective and hence the important developments in literature analysis for scientometrics is discussed next.

Literature analysis is an essential part of research, irrespective of its nature, kind, the theme addressed and the domain (or domains) in which particular research probe can be attributed to. One of the early use of scholarly literature as a means to assess scientific productivity (Lotka 1926) can be accounted to Lotka. After the upheaval of the phenomenon of ‘knowledge explosion’ and its further intensification due to digital revolution, difficulties were faced for tracking developments in scientific fields, and the modern age of scientometrics began in late 1950s. Eminent researchers like Price, Garfield, Nalimov, etc., addressed these challenges. Network approach for bibliometrics and scientometrics also initiated around that period with papers such as de Solla Price (1965) and Garfield et al. (1964). Batagelj (2003), Garfield et al. (2006), Hirsch (2005), Hummon and Doreian (1989), Rafols et al. (2010) and Waltman et al. (2010) are some among the works that contributed to the growth of the trio fields with many shared interests- bibliometrics, informetrics and scientometrics (Hood and Wilson 2001). Scholarly literature mining using citation network analysis and its applications can be found in Batagelj and Cerinšek (2013), Kajikawa et al. (2007) and Kejžar et al. (2010). Detection of paradigm shifts and emerging fields in the citation networks has been discussed in Prabhakaran et al. (2015), where Flow Vergence (FV) model was introduced. FV model is helpful for the analysts to identify the dominant mode of knowledge propagation of an article or the cluster, during a particular time point. Citation network analysis is evolving and is gaining importance as a research field. It is being looked upon as a smart methodology for extracting relevant articles from the ever increasing body of literature. State of the art of citation network analysis in the field of nanotechnology is discussed next.

In the case of citation network analysis of nanotechnology, considerable volume of publications has been accumulated over the last two decades. However, most of them focuses on nanotechnology patent literature. Huang et al. (2003) analysed nanotechnology patents between 1976 and 2002 in USPTO (United States Patents and Trademark Office). Along with content map analysis, citation network analysis was also performed. Country citation network, Institution citation network, Technology field citation network etc. were included in the citation network analysis. From technology field citation network analysis, they identified that pharmaceutical field and semiconductor devices field were exhibiting high growth during that period. In Huang et al. (2004), they updated the results for 2003 and new trends were reported. While pharmaceutical field and related chemical industry fields remained as fast growing fields, semiconductor devices field witnessed a substantial improvement in growth. Several newly formed fields were also identified in that paper. Li et al. (2007a) analysed the USPTO dataset over the period 1976–2004, by laying more emphasis on citation network analysis using network concepts and topological measures. Apart from identifying key influential players and subfields, the overall knowledge transfer patterns and knowledge transfer efficiency were also analysed. Li et al. (2007b) which shares most of the co-authors with Li et al. (2007a), compared nanotechnology patent growth in major databases like USPTO, EPO and JPO. It was found that, while USPTO dataset showed clear pattern of knowledge diffusion from highly cited fields to low cited fields, knowledge exchange was mainly between highly cited fields in the case of EPO.

Most of the works related to scholarly network analysis of nanotechnology have dealt with journal-journal citations. Leydesdorff conducted an interdisciplinary mapping of sciences for the case of nanotechnology journals in Leydesdorff (2007b) and discussed how useful it can be for identifying policy priority areas. In Leydesdorff (2007a), he analysed the usefulness of the network metric betweenness centrality (Freeman 1979) to identify the interdisciplinarity of scientific journals. Betweenness measure (normalised) identified the Journal of Nanoscience and Nanotechnology as a better interdisciplinary journal than Nano letters. Co-citation networks from scientific publications has been put to use in Avila-Robinson and Miyazaki (2013) for understanding the evolutionary paths of change of emerging nanotechnology innovation systems. Despite the developments in citation network analysis techniques and known potential of paper citation networks to unlock multitude of information for various purposes, very few articles are found to be published with the paper citation network analysis of nanotechnology publications. As nanotechnology is about to invade many techno-scientific fields of research and revolutionize many industrial practices, it is high time to extract the information about many co-existing paradigms and determine the competing ones. Lathabai et al. (2015) introduced a method to identify the pivot paper(s) of a paradigm shift, but analysis was done only on subnetworks like main path and critical path. The same is adopted here for the analysis of whole network of nanotechnology in the research area ‘engineering’. Methodology used for this work is described in the next section. Though both patent citation networks and scientific citation networks can be used for this purpose, currently we focus on scientific citation networks. Analysis of patent citation networks for the same purpose using the framework developed in this work will be one of our future endeavours.

Methodology

The schematic diagram for the analysis is shown in Fig. 1.

Fig. 1

Schema of methodology of network analysis using FV gradient

As depicted in the schema, keyword search is used to collect data from Web of Science (WoS) (Thomson Scientific et al. 2007). This type of search would bring too many hits. One can select the time span of collection and many other options to refine the search. Time span chosen was from January 1st 1989 to 22nd December 2014. As we were interested in the research area ‘engineering’, we used this research area as a filter to refine the search. Thus, the first two filters applied were done while collecting the data from WoS itself. These can be viewed as preliminary filters. In case of very huge datasets or according to the objectives of the analyst many other filter options like country, institution etc. are available. As our objectives were achieved we applied only two filters. Another filter was deployed after creating the network using histcite$^{TM}$  (Garfield et al. 2006). This is termed in Lathabai et al. (2015) as the secondary filter or network filter. The notion of k-core (Seidman 1983) with $\text {k}=1$, made the third filter. As our data processing involved the usage of three filters, the name ‘three level filter system’ was used in the schema.

In the minimal core network, we have to extract the pivot papers of paradigm shifts. The key concepts used for the extraction of pivot papers in citation networks can be found in Lathabai et al. (2015) and Prabhakaran et al. (2015). A short description of the same is given here.

Knowledge flow among published articles play a decisive role in steering the direction or course of scientific activities. Flow vergence (divergence or convergence) is an important characteristic of the flow of knowledge. Based on the dominance of flow towards it or from out of it, works can be considered to be either one of the flow vergence states—flow convergence mode or flow divergence mode. Existing network measures are not capable of identifying the direction of dominance of knowledge flow. However, as indegree is the number of papers that cited a particular paper i, it reflects the outflow of knowledge. Similarly, outdegree is the number of papers that are cited by a particular paper i, reflects the inflow of knowledge. A clear dominance of flow vergence can be identified as the difference between outflow of knowledge and inflow of knowledge. Thus difference in indegree and outdegree helps to identify the dominant mode of flow. If indegree outnumbers outdegree, then the work is surely in flow divergence mode and if the outdegree outnumbers indegree, then work is in flow convergence mode. This can be better represented in a [− 1,1] scale by using finding the ratio of the difference by the degree of the work, named as degrees ratio of the flow vergence, as shown in Eq. 1

$$\begin{aligned} W_{dr_{(fv)_i}}= \frac{indeg_i- outdeg_i}{indeg_i+outdeg_i} \end{aligned}$$

(1)

indeg and outdeg are indegree and outdegree respectively. Now if $W_{dr_{(fv)_i}} > 0$, the work is a flow divergent one and otherwise it is a flow convergent work. Thus, $W_{dr_{(fv)}}$ can be treated as an apparent indicator of ‘flow vergence potential’. Flow vergence potential can be defined as the potential for a work to exhibit higher flow vergence. As papers with $W_{dr_{(fv)}}>$ 0, is guaranteed to be in flow divergence mode, these are to exhibit high flow vergence in future too. In case of $W_{dr_{(fv)_i}}=0$, it is apparently a balanced flow of knowledge. However, as indeg of work can grow unlike the outdeg, works that already achieved a balance in present (apparent) flow vergence can exhibit high flow vergence in future (may be immediate future itself). So is the case for those works whose $W_{dr_{(fv)}}<$ 0, but are close to achieving a balance. Thus, an important question that arises is “How much close to zero can be used to distinguish papers as papers with divergence potential ?” Additionally, works with $outdeg = 0$, cannot be distinguished for their apparent flow vergence as they all have $W_{dr_{(fv)}}= 1$. This means for instance, if two papers have $indeg=30, outdeg=0$ and $indeg=1, outdeg=0$, then both will have $W_{dr_{(fv)}}= 1$. In such cases, the analysts cannot decide on the degree of importance of these works. All these signals the need for another measure, preferably a global one to indicate the relative importance in terms of quality. Such a metric exists in the network literature namely eigenvector, denoted as eig throughout this paper. On investigating whether any basic operations such as addition, multiplication, division and subtraction on the combination of $W_{dr_{(fv)}}$ and eig may yield distinction among the works that are of equal apparent potential or could raise the close to zero (balanced) and balanced flow vergence papers to divergence mode, addition is found to be suitable one. This also follows the observation in many citation networks that papers with $W_{dr_{(fv)}} < 0$ and $eig > \parallel W_{dr_{(fv)}} \parallel$ exhibit better performance in terms of their reception (indegree) and quality of reception (eigenvector), and hence by adding these net result will be a positive one. This implies an outweighing of inflow of knowledge by outflow of knowledge even if there is no apparent outnumbering of inflow by the inflow. Hence Eq. 1 can be modified to a better design of flow vergence potential as:

$$\begin{aligned} W_{FV_{i}}= \frac{indeg_i- outdeg_i}{indeg_i+outdeg_i}+eig_i \end{aligned}$$

(2)

Now, a work can be said to be of high flow vergence potential if its $W_{FV}$ is high. It is said to be in flow divergence mode if $W_{FV} > 0$.

Flow vergence gradient or FV gradient

Flow vergence index or FV indices help to identify the potential of works in a citation network. During the publication of a document as it cites some of the previous works in the network, a flow of information is supposedly occurred in the network. Associated with the flow, a potential difference start to exist between the citing-cited pair. This difference in FV potential is termed as Flow Vergence gradient or FV gradient. For an i, j pair in which j cites i, this can be expressed as:

$$\begin{aligned} \varDelta _{FV_{ij}}= W_{FV_i}-W_{FV_j} \end{aligned}$$

(3)

An important observation in citation networks is that during the time of publication and till a paper gains any citation, its FV index value will be − 1. This will be less than the papers it cited. Thus potential difference between a citing-cited pair of papers will be usually positive unless the citing paper in turn receives so many citations or citations from much good quality works. That means flow of information is mostly observed as from high potential work to a low potential work unless a not so ordinary event of reverse reset happens. When the citing paper gains much higher value of FV potential due to its merit, the flow is observed as to have occurred from a low potential work to a high potential work. This phenomenon is termed in Lathabai et al. (2015) as Flow Vergence effect or FV effect. This can be reflected by Eq. 3 as $\varDelta _{FV_{ij}}$ will take a negative value. Once the FV gradient of all arcs are computed and given as arc weights, the resulting network will be a weighted signed network.

In case of an FV effect, among a pair of papers, the citing paper is treated as the pivot paper of a paradigm shift according to Lathabai et al. (2015). Due to this property, FV gradient can be used as a paradigm shift detector. In Lathabai et al. (2015), the use of FV gradient analysis was conducted only on subnetworks such as main path and critical path of the field ‘Biotechnology for engineering’ and the entire network was not analysed. Due to the immense possibilities offered by an extensive analysis of FV gradient on the entire network, in this work we have done the same. Since the arcs with negative weights capture the flow vergence effect, the pairs of papers that form negative FV gradients can be extracted along with such arcs. Once this is achieved, one could proceed to analyse the content of these papers so as to identify co-existing, competing paradigms or complementing paradigms. Also the clusters which are extended subnetworks (to the immediate neighbourhood) of the subnetworks formed by pairs of papers could be analysed to derive the structural implications of co-existing paradigms which would lead to give confirmatory insights about whether they are competing or complementary etc. Some thumb rules for the identification of nature of these paradigms are introduced here as the Dichromatic framework.

Dichromatic decision framework for identification of nature of paradigm

As FV gradient is an arc metric, it gives pairs of papers that are instrumental in paradigm shift. Here, the source paper of the arc (the one which receive knowledge) is the pivot paper of paradigm shift. The target paper of the arc (the one which passed knowledge to the pivot) is more likely to represent a development from the previous paradigm than the initial developments in the new paradigm. Dichromatic decision framework, as the name suggests, uses two different colours to distinguish the pivot papers from its flow vergence pair. In this paper, we use red colour to represent the pivot papers and blue colour to represent its pair. For the analysis of extended subnetworks (clusters), dichromatic decision framework lays some thumb rules which may help to deduce the nature of the co-existing paradigms. These rules are given as follows:

Fig. 2

Possible situations where Dichromatic decision framework can be used

• 1. If there is a direct knowledge transfer from red paper of one paradigm to the red paper of another paradigm, then these paradigms might be complementary. Here knowledge transfer enhances the paradigm that received knowledge and the receiver paradigm do not offer a threat of competition of ‘takeover’ to the source paradigm. This situation is illustrated in Fig. 2a.

• 2. If there is a direct knowledge transfer from red paper of one paradigm to the blue paper of another paradigm, then these paradigms might be competing. In this case, blue paper of one paradigm receives knowledge from red paper of another. This particular blue paper may be dealing with an incremental contribution that is in someway related to the one addressed in red paper of the other paradigm. Here as there is a high chance for theme addressed in red paper to be much advanced than the blue one, the contribution of the red paper which is the FV pair of blue one that received knowledge from another red paper can impede the progress of theme in cited red paper. This is illustrated in Fig. 2b.

• 3. If there is a direct knowledge transfer from blue paper of one paradigm to the blue paper of another paradigm, then these paradigms can be complementary or merely co-existing. Here there is no visible signs of competition. This is depicted in Fig. 2c.

• 4. If there is a direct knowledge transfer from blue paper of one paradigm to the red paper of another paradigm, then these paradigms can be competing. Here this can be viewed as branching paradigms that usually competes for establishment with dominance over other. This can be seen in Fig. 2d.

• 5. If there is no direct knowledge transfer, but the two paradigms are some how appearing in the same cluster (due to knowledge transfer through papers that are not part of any of the FV pairs), then these are independently developing paradigms that co-exists for time being, but can be complementary to each other in future. This is shown in Fig. 2e.

• 6. If there is only one paradigm in an extended subnetwork, then this is an independently developing paradigm that co-exist with all other paradigms. It may complement any of the paradigms in other clusters in future. Illustration of the independent co-existence can be found in Fig. 2f.

Results and discussions

The details of data used in this paper is mentioned first. After that, analysis of the resulting subnetworks, its results and findings are given. Implications of the findings are given in the last part of this section.

Data description

Data collected from WoS using the keyword ‘Nanotechnology’ during the time span January 1st 1989 to December 22nd 2014, filtered by research area ‘engineering’ returned almost 2778 hits. After creating the citation network, third filter (k-core with $\text {k}=1$) was applied to remove the isolated papers and then the network had 1076 papers and 1544 arcs (citation links). The minimal core of the network (filtered network using $\text {k}=1$) is shown in Fig. 3. This network is subjected to the FV gradient analysis to determine the pivot papers of paradigm shifts.

Fig. 3

Minimal core of nanotechnology for engineering

FV gradient analysis

Fig. 4

Left 18 subnetworks formed due to flow vergence effect, Right 10 clusters formed due to the extension of 18 subnetworks

In the network shown in Fig. 3, FV indices of all the works are computed using Eq. 2. Then FV gradients of all the arcs are computed using Eq. 3. After that the papers associated with the arcs with negative weights (as these arcs reflect the flow vergence effect) along with the arcs are extracted from the network. 25 arcs i.e., 25 pairs of papers were obtained. They formed 18 small subnetworks as shown in Fig. 4 (left). On extension to their immediate neighbourhood, these subnetworks were found to form 10 clusters. These are shown in Fig. 4 (right). Now, we are ready for a detailed analysis of the contents of the 18 subnetworks. First 3 subnetworks are shown in Fig. 5.

Fig. 5

First 3 subnetworks (left to right) with pivot papers shown in red. (Color figure online)

Subnetwork 1

Subnetwork 1, shown in Fig. 5 left, consists of three papers. First one, 195 Bohr MT, 2002 is titled as Nanotechnology goals and challenges for electronic applications. This paper lists out and describes the basic devices such as Carbon nanotube Field Effect Transistors, nanowire FETs, single electron transistors and molecular devices and circuit requirements for electronic logic and memory products, which should be considered while evaluating nanotechnology options as future replacements for silicon MOSFETs (Metal Oxide Semiconductor Field Effect Transistors). This paper was cited by 487 Cao Y, 2004 and 488 Frechette MF, 2004. The title of the former is The future of nanodielectrics in the electrical power industry and that of the latter is Introductory remarks on nanodielectrics. While the former was published from US, the latter was from Canada. Both the papers discuss about nanodielectrics. In former, Cao et al. discussed about the short and long term future R&D needs from the point of view of industrial applications. Frechette et al., in latter, based their discussion on the impact of neologism nanotechnology (Frechette et al. 2004) on materials such as dielectrics and their insulating properties. The paper was concluded by introducing a tentative definition of ‘nanodielectrics’. Both 487 Cao Y, 2004 and 488 Frechette MF, 2004 were identified as pivot papers of paradigm shifts and are shown red in Fig. 5.

Both these independent works, carried out during the same period and from two different countries, have contributed to the development of ‘nanodielectrics’ as a field of research and hence these can be rightly considered as the pivot papers of the paradigm shift towards ‘nanodielectrics’ from dielectrics.

Subnetwork 2

334 John DL, 2003 is the first paper in the subnetwork 2. It is titled as Electrostatics of coaxial Schottky-barrier nanotube field-effect transistors. Unlike ordinary FETs, an unusual possibility of simultaneous contributions of electrons and holes to the drain current was revealed for the nanotube FETs. This paper is found to be cited by the first pivot paper in the subnetwork, 434 Clifford JP, 2004. Titled as Electrostatics of partially gated carbon nanotube FETs, it investigated various electrostatic properties of partially gated co-axial carbon nanotube FETs. It was found that the introduction of charge is found to be effective in suppressing the dominant role played by the end contacts in the barrier to the charge flow and hence offered the possibility of bulk control. This paper is inturn cited by 647 Hanson GW, 2005. It is titled as Fundamental transmitting properties of carbon nanotube antennas and as the name suggests, the properties of carbon nanotubes as antenna elements were investigated.

Contributions of both 434 Clifford JP, 2004 and 647 Hanson GW, 2005 were vital in unveiling the properties of carbon nanotubes that can be applied for communication purposes. A paradigm shift towards ‘nanoelectronics based on Carbon NanoTubes (CNTs)’ from ordinary electronic devices, especially on antenna design, were identified and this may have an impact on the communication systems.

Thus it is evident that, the themes addressed by subnetworks 1 and 2 are related to communication technologies and thus the paradigms discussed in the pivot papers are co-existing paradigms. This can be considered as an evidence of the ability of FV gradient analysis to determine the co-existing paradigms.

Subnetwork 3

352 Chang CY, 2003 is one of the two papers that made its entry into the third subnetwork formed on FV gradient analysis. It is titled as The highlights in the nano world, and was published in IEEE proceedings. Successful realization of MRAM (Magnetic RAM) by using nanotechnology and spintronics (Pseudo Spin Valve technology) was demonstrated. Various potential emergent applications of nanobiotechnology such as fabrication of nanodevices using biotechnology and nanobiomedical applications of nanoparticles were discussed as well. The second paper in the subnetwork is 1253 Akyildiz IF, 2008, titled as Nanonetworks: a new communication paradigm. The title itself signifies the importance of this paper as a pivot and can be viewed as another testimony for the usefulness of FV gradient analysis in the detection of pivot papers of a paradigm shift. In this tide turning paper, nanonetworks or the interconnection of nano-machines were discussed.

Nano machine is a tiny component, an arranged set of molecules, that could perform only very simple tasks. Nanonetworks are supposed to expand the capabilities of standalone nano-machines to cooperate and share information. For the establishment of this paradigm, development of network components and molecular communication theory that enable the use of molecules to encode and transmit information instead of electromagnetic or acoustic waves, novel solutions such as development of new architecture and protocols, etc., are the open research challenges that have to be solved (Akyildiz et al. 2008). According to the authors (Akyildiz IF et al.), if all these could be addressed successfully, nanotechnology deployment can be anticipated before 2030. Considering the present intensity of R&D activities undertaken by various research institutions all over the world, this vision seems to be achievable and we may witness the realization of Internet of Nano Things (IoNT) (Akyildiz and Jornet 2010)—a cross road of BT, IT and Nanotechnology, by then. Subnetworks 4 to 6 are shown in Fig. 6.

Fig. 6

Fourth, Fifth and Sixth subnetworks, with pivot papers shown in red. (Color figure online)

Subnetwork 4

The paper labelled as 851 Hanson GW, 2006 and titled as On the applicability of the surface impedance integral equation for optical and near infrared copper dipole antennas forms the first paper of this subnetwork. This paper was single handedly authored by Hanson GW and published in journal- IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION. Hanson is a notable name as the second pivot paper in the second subnetwork was already credited to his account. The pivot paper in this subnetwork is also written by Hanson and is having the label- 1136 Hanson GW, 2008 and title- Dyadic Green’s functions for an anisotropic, non-local model of biased graphene. On investigation with varying the bias, it was revealed that one can exert significant control over graphene’s electromagnetic propagation characteristics, including guided surface wave phenomena. This may be vital for future electronic and optical communications. This can be treated as a pivot of paradigm shift towards graphene based electronic and optical communication.

Though subnetworks 2 and 4 deal with antenna and wave propagation design for communication, and nanotube antennas were discussed in subnetwork 2, subnetwork 4 elaborates on the properties of graphene and its potential applications in communication. Thus, these two paradigms can be viewed as competing paradigms.

Subnetwork 5

913 Lee SH, 2007 is the label of the first paper in the subnetwork 5. Its title is Nanoparticles of poly(lactide)-Tocopheryl polyethylene glycol succinate (PLA-TPGS) copolymers for protein drug delivery. Upon several tests including the test for structural integrity of the released proteins and its durability of retention, NPs (nanoparticles) of PLA:TPGS was found to provide the encapsulated protein a milder environment and hence such NPs could be of great potential for the clinical formulation of proteins and peptides (Lee et al. 2007). 1746 Gan CW, 2010 is the second paper of this subnetwork, as well as the pivot paper of the paradigm shift in nano particle based targeted drug delivery. It is titled as Transferrin-conjugated nanoparticles of Poly(lactide)-D-alpha-Tocopheryl polyethylene glycol succinate diblock copolymer for targeted drug delivery across the blood-brain barrier. This paper, which share the co-author Feng SS with 913 Lee SH, 2007, discussed first about the development of a nanoparticle system for targeted drug delivery across the blood brain barrier (BBB). Experiments showed that transferrin (Tf)-conjugated PLA-TPGS NPs formulation of Docetaxel could be more efficient than the other formulations after 24 h treatment. Though not satisfactory, preliminary investigation (ex vivo biodistribution) also revealed that the Tf-conjugated PLA-TPGS NPs formulation could deliver imaging/therapeutic agents across the BBB (Gan and Feng 2010).

Targeted drug delivery is one of the subfield that offer great room for the expansion of nanotechnology. Unlike the previously discussed subnetworks, in which nanotechnology invades the subfields occupied by conventional technologies, the targeted drug delivery subfield witnesses expansion of nanotechnology occupied landscape. Thus competitive paradigms can be expected in this subfield and exploration of other subnetworks proved us right.

Subnetwork 6

This subnetwork consists of two papers. First among them is 923 Bhikkaji B, 2007, titled as PVPF control of piezoelectric tube scanners. This work deals with the nanotechnology applications in topographical mapping of material surfaces. Since speed and precision are vital for obtaining good topographical maps, a control system- Positive Velocity Position Feedback (PVPF) controller is introduced in order to damp down the unwanted resonant modes. Experimental results confirms a good tracking performance due to the dampening of these troublesome resonant modes (Bhikkaji et al. 2007). The paper that drew knowledge from this paper is 998 Devasia S, 2007, titled as A survey of control issues in nanopositioning. While the previous paper insisted on the need for a controller to maintain the speed for positioning in topographical mapping, this paper, emphasises the key attributes and critical factors required for the successful design of a nanopositioner (precision mechatronic systems) (Devasia et al. 2007) in general. An overview of the nanopositioning techniques and devices available then was given, along with an emphasis on the role of advanced control techniques for ensuring the accuracy and speed of operation of these systems. This emphasis on the need for feedback control for different devices might have inspired further design improvements and this paper can be regarded as a pivot paper in paradigm shift from ordinary nanopositioning systems towards ‘precise nanopositioning systems with control systems’. Subnetworks 7 to 9 are shown in Fig. 7.

Fig. 7

Seventh, Eighth and Ninth subnetworks, with pivot papers shown in red. (Color figure online)

Subnetwork 7

This subnetwork is shown in Fig. 7 (leftmost). The first paper in this subnetwork is having a label 937 Cho H, 2007 and is titled as Adder designs and analyses for quantum-dot cellular automata. This paper deals with the basic building block of nanotechnology circuits, named as ‘Quantam dot cells’. Gates, wires, and memories etc. can be made using these. According to Cho and Swartzlander, quantum dot cellular automata (QCA) is an emerging nanotechnology for electronic circuits, with the advantages such as increased speed, smaller size and low power consumption. Three kinds of adder (Ripple carry adders, carry lookahead adders, conditional sum adders) designs in QCA are introduced in that paper for larger circuits and designs are simulated with different operand sizes. Results are also compared according to complexity, area and design (Cho and Swartzlander 2007). The second paper in this subnetwork, is also co-authored by Cho and Swartzlander, and can be viewed as the natural extension of the first one. Its label is 1443 Cho H, 2009 and is titled as Adder and Multiplier Design in Quantum-Dot Cellular Automata. This paper deals with the design of a carry flow adder using unique QCA characteristics and simulations indicate good performance in terms of complexity, area and delay. Designs of Serial parallel multipliers were also explored and Serial parallel multiplier designs were simulated using several operand sizes (Cho and Swartzlander 2009). Thus this paper made significant contributions to the new paradigm ‘Quantum dot cellular automata’ which could enhance the shift from traditional (transistor based technology) electronic circuit design to ‘nanotechnology based electronic circuit design’. Owing to the applications of electronic circuits in other fields, this co-existing paradigm can complement many nanotechnology paradigms in future.

Subnetwork 8

1082 Ostertag K, 2008 is the label of first paper in this subnetwork. Its title is Identification of starting points for exposure assessment in the post-use phase of nanomaterial-containing products. As suggested by the title, this paper deals with nanomaterial-containing products and a pragmatic methodology was proposed for “the systematic identification of knowledge gaps and of priority starting points for exposure assessment as one step of a risk assessment” (Ostertag and Hüsing 2008). A demonstration of the usefulness and feasibility of the methodology on post use phases of automobiles and paper was included. It is envisaged that the methodology once transferred to other life cycle phases could be useful for selected nanoparticles or products that are still under development. Second paper of the subnetwork, labelled as 1383 Meyer DE, 2009 and titled as An Examination of Existing Data for the Industrial Manufacture and Use of Nanocomponents and Their Role in the Life Cycle Impact of Nanoproducts, examined the manufacture and use of nanocomponents and their possible effects on the Life Cycle Impact (LCI) of resulting nanoproducts. Based on the available data of nanoproducts and nanocomponents, they first identified major groups of nanocomponents such as inorganic nanoparticles, carbon-based nanomaterials, and specialty/composite materials and studied their role in LCI of nanoproducts (Meyer et al. 2009). While the first paper focused mainly on the post-use phase of products, the second seemed to deal with the full life cycle of such products. This can be regarded as a pivot paper in paradigm shift towards the life cycle assessment of nanoproducts using their nanocomponents, i.e., a shift towards life cycle assessment methodology for nanotechnology products from the traditional life cycle assessment methodologies used for ordinary (non-nano) products.

This paradigm is related to manufacture and may complement other nanotechnology paradigms in future, due to the potential spread and reach of its applications.

Subnetwork 9

First paper in this subnetwork is 1125 Rachlin E, 2008 and it is titled as Analysis of mask-based nanowire decoders. This paper identified the major nanotechnology challenge in controlling parallel sets of nanowires (NW) using only moderate number of mesoscale wires, in case of stochastically assembled nanoscale architectures that could achieve 100 times greater density than achieved by CMOS devices. Out of the three available methods: (1) method based on NW differentiation during manufacture, (2) method of making random connections between NWs and mesoscale wires, (3) mask-based approach (Rachlin and Savage 2008)—a method of interposition of high-k dielectric between NWs and mesoscale wires, mask based approach has been analysed. It was found that this method requires a large number of mesoscale control wires for its realization compared to other two schemes. Second paper in this subnetwork is Dai JW, 2009 and titled as Analysis of Defect Tolerance in Molecular Crossbar Electronics. In that paper, major obstacles in the future ‘nanocomputing paradigm’ (i.e., computation using molecular electronics based on NWs and CNTs) were identified and they are ‘Excessive defects from bottom up stochastic assembly’. An information theoretic approach was presented to “investigate the intrinsic relationship between defect tolerance and inherence redundancy in molecular crossbar systems” (Dai et al. 2009). Through this method, they derived the gap of reliability between redundancy-based defect tolerance, ideal defect-free molecular systems and also given implications to the related design optimization problem.

The second paper undoubtedly qualifies as a pivot paper of paradigm shift related to ‘nanotechnology for integration’ which could be crucial for the establishment of ‘nano computing’ paradigm. Thus, this new paradigm detected by FV gradient analysis is also a complementary paradigm which can leverage many other paradigms in future. Subnetworks 10 to 12 are shown in Fig. 8.

Fig. 8

Tenth, Eleventh and Twelfth subnetworks, with pivot papers shown in red. (Color figure online)

Subnetwork 10

Targeted delivery of paclitaxel using folate-decorated poly(lactide)-vitamin E TPGS nanoparticles is the title of first paper 1190 Pan J, 2008 in the tenth subnetwork. In this paper, it was shown that the NP formulation outperformed pristine drug in better therapeutic effect and had many advantages. Experiments were conducted using Paclitaxel as model drug (Pan and Feng 2008). From the analysis of first paper itself, it is clear that this subnetwork also deals with the use of ‘nanoparticles in targeted drug delivery’. Titled as Targeting and imaging cancer cells by Folate-decorated, quantum dots (QDs)-loaded nanoparticles of biodegradable polymers and labelled as 1362 Pan J, 2009, pivot paper in this subnetwork presented a new strategy to prepare folate-decorated NPs for Quantum Dot (QD) formulation for targeted and sustained imaging of cancer cells (Pan and Feng 2009). This can be regarded as a paradigm shift for cancer detection (using nanoparticle based imaging) over conventional imaging techniques. This QD based imaging is regarded as a complementary paradigm for ‘targeted drug delivery’.

Subnetwork 11

The first paper (1193 Hanson GW, 2008) in this subnetwork deals with nano-radius dipole antennas. It is titled as Radiation efficiency of nano-radius dipole antennas in the microwave and far-infrared regimes. Since the radiation efficiency of nano-radius antennas is found low in microwave and far-infrared frequencies, several methods to improve the same are discussed. These methods include the use of superconducting nanowires and multi-wall carbon nanotubes (Hanson 2008). Second paper, 1825 Hanson GW, 2011, titled as A Common Electromagnetic Framework for Carbon Nanotubes and Solid Nanowires-Spatially Dispersive Conductivity, Generalized Ohm’s Law, Distributed Impedance, and Transmission Line Model. In this paper, various propagation constants of nanowires and nanotubes are numerically presented along with general formulation of spatially dispersive conductivity, distributed impedance, Ohm’s law relation, and transmission line model of both carbon nanotubes (CNTs) and solid material nanowires (Hanson 2011). This can be regarded as a shift from conventional electromagentic framework for communication towards the ‘nanotechnology based electromagnetic framework for communication’. The establishment of the same could complement the likely to be established paradigm addressed in second subnetwork.

Subnetwork 12

The first paper, 1264 Godman M, 2008, in this subnetwork, titled as But is it unique to nanotechnology? was published in SCIENCE AND ENGINEERING ETHICS. Godman attempted to establish nanoethics as a new sub-discipline of applied ethics. The nature of this sub-discipline was discussed and concluded that the conditions for nanoethical debate differ according to the nature of values (whether the values are internal to the technological development or external to it) (Godman 2008) in question. Internal values can be the kinds like health and safety and external issues include privacy, equity etc. 2057 Bjornstad DJ, 2011 is the pivot paper identified from the subnetwork 12, which is also published in SCIENCE AND ENGINEERING ETHICS. This paper is titled as Adding to the Mix: Integrating ELSI into a National Nanoscale Science and Technology Center. Issues associated with integrating the study of Ethical, Legal and Social Issues (ELSI) into ongoing scientific and technical research were discussed and an approach adopted by the authors for their own work with the center for nanophase materials sciences (CNMS) at the Oak Ridge national laboratory (ORNL) was described. It was found that “Successful integration is dependent on the particular ELSI question set that drives the project” (Bjornstad and Wolfe 2011). This paper can be regarded as a pivot paper in paradigm shift towards ‘nanoethics’ which demands the integration of ‘ELSI’ into science and engineering ethics.

By nature this shift has the potential to impact and leverage all the R&D efforts in every other upcoming paradigms of nanotechnology and hence this could be a complementary paradigm once nanotechnology paradigms begin to take over the techno-economic development. Subnetworks 13 to 15 are shown in Fig. 9.

Fig. 9

Thirteenth, Fourteenth and Fifteenth subnetworks, with pivot papers shown in red. (Color figure online)

Subnetwork 13

In this subnetwork, the first paper, 1430 Choi JY, 2009 is titled as The Impact of Toxicity Testing Costs on Nanomaterial Regulation. A very important theme is addressed in this paper- toxicity of nanoparticles. The information about toxicity is crucial for the regulation of the use of nanoparticles. It was estimated that if all the then existing nanomaterials with U.S. were to be thoroughly tested for toxicity, it would take 34–53 years. A tiered risk-assessment strategy similar to the EU’s REACH legislation for regulating toxic chemicals (Choi et al. 2009) is suggested. Second paper, 1504 Godwin HA, 2009, titled as The University of California Center for the Environmental Implications of Nanotechnology, deals with almost all the environmental aspects of nanotechnology as addressed by the UCCEIN (University of California Center for the Environmental Implications of Nanotechnology). Seven interdisciplinary research groups were formed in the center for directing regulatory mechanisms towards predictive toxicology and safe design of nanomaterials (Godwin et al. 2009). This paper clearly played a pivotal role towards the establishment of a new paradigm with ‘strong scientific foundation for safe implementation of nanotechnology, for innovative biological and environmental applications’.

Like ‘nanoethics’, the paradigm- ‘nanoenvironmental impact regulation’ that could redefine the environmental impact regulation of research, is likely to complement the future paradigms once nanotechnology invades and takes over almost all the production and manufacturing sectors.

Subnetwork 14

1509 Chun YW, 2009 is the first paper in this subnetwork. Titled as The Role of Nanomedicine in Growing Tissues, this review paper, presented the application of nanomaterials to regenerate numerous organs (specific examples are bone, neural, and bladder tissues) along with necessary future directions for the progress of the field ‘nanomedicine’ (Chun and Webster 2009). Second paper of the subnetwork, 1998 Ueno T, 2011 is titled as Enhanced bone-integration capability of alkali- and heat-treated nanopolymorphic titanium in micro-to-nanoscale hierarchy. This paper introduced nanopolymorphic features of alkali-and heat-treated titanium surfaces and determined how the addition of these nanofeatures to a microroughened titanium surface affects bone implant integration. Acceleration and elevation of bone implant integration due to Nanofeature-enhanced osteoconductivity, had been clearly demonstrated (Ueno et al. 2011). Third paper of the subnetwork, 2184 Rani VVD, 2012 is titled as Osteointegration of titanium implant is sensitive to specific nanostructure morphology. Nanostructured implant surfaces are known to enhance osteoblast activity, which is vital for orthopedic implant integration. It was shown that “a specific surface nanomorphology, viz. nanoleaves, which is a network of vertically aligned, non-periodic, leaf-like structures with thickness in the nanoscale, provided a distinct increase in osteoblast cell proliferation, alkaline phosphatase (ALP) activity and collagen synthesis compared to several other types of nanomorphology, such as nanotubes, nanoscaffold and nanoneedles (rods)” (Rani et al. 2012).

Since second paper analysed the effects of nanopolymorphic features on titanium surfaces on bone implant integration and the third paper identified the surface nanomorphology that enhanced the activities vital for bone implant integration, these papers can be rightly regarded as pivot papers of paradigm shifts towards ‘nanotechnology based bone implantation’ from conventional orthopedic implantation. This comes under a major shift towards ‘nanotechnology based artificial organ implantation’ from ‘conventional organ implantation’.

Subnetwork 15

This subnetwork consists of 4 papers, which include 3 pivot papers of paradigm shift. 1591 Liu YT, 2010 is the first paper in the subnetwork, titled as Folic acid conjugated nanoparticles of mixed lipid monolayer shell and biodegradable polymer core for targeted delivery of Docetaxel. It discussed about the development of a system of nanoparticles of mixed lipid monolayer shell and biodegradable polymer core for targeted delivery of anticancer drugs with Docetaxel as a model drug. It was proved that folic acid conjugated nanoparticles of mixed lipid monolayer shell and biodegradable polymer core could provide a drug delivery system of precise control of the targeting effect (Liu et al. 2010b). Second paper in the subnetwork (first pivot paper), 1784 Liu YT, 2010 is titled as A strategy for precision engineering of nanoparticles of biodegradable copolymers for quantitative control of targeted drug delivery. As “quantitative control of targeting effect for the drug delivery system of ligand-conjugated nanoparticles of biodegradable polymers is at the cutting edge in the design of drug delivery device” (Liu et al. 2010a), a post conjugation strategy was developed. The post-conjugation strategy was found to provide more efficient use of the ligand and protected its bioactivity in the nanoparticle preparation process, thus resulting in much better performance in drug targeting. Second pivot paper in the subnetwork, 1905 Mi Y, 2011, titled as Formulation of Docetaxel by folic acid-conjugated D-alpha-tocopheryl polyethylene glycol succinate 2000 (Vitamin E TPGS(2k)) micelles for targeted and synergistic chemotherapy. This paper deals with “a new concept in the design of drug delivery systems—the carrier materials of the drug delivery system can also have therapeutic effects, which either modulate the side effects of, or promote a synergistic interaction with the formulated drug” (Mi et al. 2011). Drugs Paclitaxel and Docetaxel were used for the study and this design of drug delivery system is competing one to the paradigm which is attempted to be established by the 1784 Liu YT, 2010. Third pivot paper which is likely to deal with another paradigm shift is 2367 Kong SD, 2013, titled as Magnetic field activated lipid-polymer hybrid nanoparticles for stimuli-responsive drug release. “Stimuli-responsive nanoparticles (SRNPs) offer the potential of enhancing the therapeutic efficacy and minimizing the side-effects of chemotherapeutics by controllably releasing the encapsulated drug at the target site” (Kong et al. 2013). A lipid-polymer hybrid nanoparticle system containing magnetic beads for stimuli-responsive drug release using a remote radio frequency (RF) magnetic field was reported. This scheme has the strengths such as ease of preparation, stability, and controllable drug release and provide the opportunity to improve cancer chemotherapy. Again, this is a competing paradigm with the ones addressed by first and second pivot papers. Paradigm attempted to be established by the first pivot is complementary to the ones addressed in other subnetworks which deals with targeted drug delivery. Subnetworks 16 to 18 are shown in Fig. 10.

Fig. 10

Sixteenth, Seventeenth and Eighteenth subnetworks, with pivot papers shown in red. (Color figure online)

Subnetwork 16

1602 Nikulainen T, 2010 is the first paper in this subnetwork and it is titled as Transferring science-based technologies to industry-Does nanotechnology make a difference? This paper attempted to investigate “whether there is a need for nanotechnology-specific policies to facilitate nanotechnology transfer from universities to firms” (Nikulainen and Palmberg 2010), using individual-level survey data of university researchers in the Finnish nanotechnology community. It was found out that, active university researchers in nanotechnology were endowed with motivations and show interactions with industry unlike researchers in other disciplines. However, challenges faced by them were altogether different from those in other disciplines. Hence, a need for ‘nanotech-specific’ policies was stressed. 1850 Allarakhia M, 2011 is the second paper in the subnetwork, titled as Managing knowledge assets under conditions of radical change: the case of the pharmaceutical industry. This paper discussed about a major paradigm change in pharmaceutical industry, in which, advances in biology, nanotechnology, and the computational sciences were found to drive the technological developments. The effect of this change in intellectual property (IF) management practices were probed. In order to assist firms to effectively manage both knowledge assets and associated intellectual property in the current paradigm, an IF model ‘the transition point model’ was designed (Allarakhia and Walsh 2011).

This shift is associated with the radical change in pharmaceutical technologies due to invasion of nanotechnology in pharmaceutical technologies. Like ‘nano ethics’ and ‘nano environmental impact assessment’, though not directly, this paradigm can become complementary paradigm to many other paradigms as it deals with knowledge management and intellectual property management of pharmaceutical firms.

Subnetwork 17

The first paper in this subnetwork is 2113 Yang XK, 2012, titled as A comparative analysis and design of quantum-dot cellular automata memory cell architecture. Though published in 2012, its online record version was available from July 2010 onwards and hence cited by the second paper which is published in 2011. The event in which a publication cites another publication which is published after it occurred here as a result of the difference in production speeds of the journals.

Drawbacks of the previous QCA memory architectures were analysed in 2113 Yang XK, 2012 and a new layout had been presented for memory architecture that exploited regular clock zone layout by employing two new clocking signals and a compact Read/Write circuit (Yang et al. 2012). 1917 Tehrani MA, 2011 is the second paper in this subnetwork. Its title is Design and implementation of Multistage Interconnection Networks using Quantum-dot Cellular Automata. This paper presented the first design methodology of Multistage Interconnection Networks (MINs) using QCA (Quantan-dot Cellular Automata). Performance evaluations for demonstrating the functionality and validity of this design methodology had been done through simulation (Tehrani et al. 2011). This can be regarded as the invasion of nanotechnology over the memory cell or storage technology as well as integration for computing. As storage is crucial for computation purposes, this shift towards nanotechnology i.e., (MIN-QDCA) based architecture of memory cell electronic circuitry could complement the ‘nano computing’ paradigms.

Subnetwork 18

This subnetwork consists of 4 papers, of which 3 are pivot papers. There are two branches: left branch with 1 pivot paper and the right branch with two pivot papers. First paper (non-pivotal paper) is 2262 Mi Y, 2012, titled as Multimodality treatment of cancer with herceptin conjugated, thermomagnetic iron oxides and docetaxel loaded nanoparticles of biodegradable polymers.A system of nanoparticles named as MultiModality NanoParticles (MMNPs) were developed with the formulation of docetaxel for chemotherapy, herceptin for biotherapy and targeting, and iron oxides (IOs) for hyperthermia therapy. It was “demonstrated that the MMNPs achieved a significantly higher therapeutic effects than the various combination of the corresponding individual modality treatment NPs and the dual modality treatment NPs due to the synergistic effects among the chemo, bio, and thermo therapies” (Mi et al. 2012).

Second paper (pivot paper in the left branch) is 2325 Bao G, 2013, which is titled as Multifunctional Nanoparticles for Drug Delivery and Molecular Imaging. Unlike the first paper, which discussed about the use of MMNPs with a particular formulation, this paper reviewed “major forms of multifunctional nanoparticles that have emerged over the past few years, and provide a perceptual vision of this important field of nanomedicine” (Bao et al. 2013). Major contribution of this paper is that it clearly discusses about a paradigm shift from the use of nanoparticles with one or two functions to the usage of multifunctional nanoparticles. This shift can be viewed as a competing paradigm to the ‘nanoparticle based drug delivery’ and ‘nanoparticle based imaging’.

Right branch consists of two papers. First one (third paper in the subnetwork) is 2466 Zeng XW, 2013, which is titled as Cholic acid-functionalized nanoparticles of star-shaped PLGA-vitamin E TPGS copolymer for docetaxel delivery to cervical cancer. In this paper, a new system of nanoparticles (NPs) of cholic acid functionalized, star-shaped block copolymer consisting of PLGA and vitamin E TPGS for sustained and controlled delivery of docetaxel for treatment of cervical cancer, had been developed (Zeng et al. 2013). Though not capable of multifunctionality, this shift from conventional drug delivery to nanoparticles of cholic acid functionalized delivery for docetaxel could offer competition to the shift towards other systems for nanoparticle based drug delivery for docetaxel, that had been discussed earlier.

Latest pivot found in this subnetwork is 2609 Zhang XD, 2014, titled as The effect of autophagy inhibitors on drug delivery using biodegradable polymer nanoparticles in cancer treatment. This has been published in 2014, which is the end time of the time span of our analysis. This indicates the capability for early detection of a paradigm shift of FV gradient method. This paper also deals with the paradigm of cholic acid conjugated NP based delivery of docetaxel. However, this paper contributed immensely to nanomedicine field by exploring the ‘fate’ of of the NPs after their internalization into the cells. It was found that “NPs induce autophagy of the cancer cells and thus may hinder the advantages of the nanomedicine” (Zhang et al. 2014). A new mechanism of “cancer cells to have PLGA NPs captured and degraded by auto-lysosomes” (Zhang et al. 2014), was also reported.

Since an anomaly has been identified in the NP based drug delivery even in the Cholic acid NP formulation, the field of nanomedicine, especially targeted drug delivery for cancer treatment requires more paradigm shifts, which might occur in the near future.

Extended subnetwork analysis

Like earlier mentioned, when the 45 pairs of papers in the 18 subnetworks were extended to their immediate neighbourhood, 10 much bigger subnetworks or clusters (with a total of 220 papers and 272 arcs) were formed. These clusters are further analysed for deriving structural implications about the existence of various co-existing paradigms and to determine whether they are of complementary or competing nature. The structural analysis of each clusters or extended subnetworks are given below:

Fig. 11

Extended subnetwork showing the first subnetwork

Cluster #1

This cluster is shown in Fig. 11. It consists of subnetwork 1 alone. As subnetwork 1 deals with an emerging paradigm ‘nanodielectrics’ it might influence other paradigms such as ‘nanocomputing’ and ‘nanoelectronics’ in due course. For now, this paradigm seems to be emerging independently.

Fig. 12

Extended subnetwork showing the subnetworks 2, 3, 4 and 11

Cluster #2

Second cluster is shown in Fig. 12. As evident from Fig. 12, it consists of 4 subnetworks 2, 3, 4 and 11. Piovot papers in each subnetwork are shown in red and the papers that are cited by pivot paper and influenced by ‘FV effect’ is shown in blue.

Fig. 13

Extended subnetwork showing the subnetworks 5, 10, 15 and 18

Table 1 Different subnetworks in cluster 2, themes and nature of paradigms and influence on other subnetworks

From the Fig. 12, it can be found that, subnetwork 2, which deals with a paradigm ‘Carbon nanotube (CNT) based nanoelectronics’ is extended further by two other subnetworks, 4 and 11. Themes represented by all the pivot papers of paradigm shifts are given in Fig. 12 and also in Table 1. Out of these, a clear competing paradigm for 2 is subnetwork 4, which deals with a paradigm ‘Graphene based nanoelectronics’. If ever CNT based electronic products become obsolete in future, it might be due to graphene based products. CNT based nanoelectronics also led to a new paradigm in ‘nano electromagnetism’. This also pose threats to the CNT based nanoelectronics as electromagnetic nano devices might have potential to replace some of the CNT based products—especially communication devices like nanoantennas. This may happen mainly because the creation of a general electromagnetic framework might accelerate R&D related to the usability of nanomaterials other than CNT for the design of antennas and communication devices.

Subnetwork 3, which deals with ‘nanonetworks’ or ‘internet of nanothings’ is an example of a co-existing paradigm. However, it has the potential to drive or determine the direction of research in all other subnetworks.

Cluster #3

Third cluster is shown in Fig. 13. Its deals with a major subtheme of biomedical applications of nanotechnology—‘targeted drug delivery’. It consists of 4 subnetworks 5, 10, 15 and 18, can be found in Fig. 13.

Table 2 Different subnetworks in cluster 3, themes and nature of paradigms and influence on other subnetworks

All the subnetworks deal with ‘nanoparticle based targeted drug delivery and/or imaging’ for cancer treatment. Themes represented by all the pivot papers of paradigm shifts are given in Fig. 13 and also in Table 2. Subnetworks paradigms in 5 and 10 do not compete with each other. Three branching paradigms found in subnetwork 15 are competitive among themselves. So do the two branching paradigms in subnetwork 18. One of the branching paradigms (branch 1) in subnetwork 15 can be treated as complementary paradigms to the one addressed in subnetwork 10. Two branching paradigms in subnetwork 18 competes with all the three paradigms of subnetwork 15 and also the one in subnetwork 10. Eventually, the two paradigms in 18 or anyone of them may offer competition to the one in subnetwork 5. As of now, the influence of these two branching paradigms over the one in subnetwork 5 is not clear.

Fig. 14

Extended subnetwork showing subnetwork 6

Cluster #4

As shown in Fig. 14, this extended subnetwork contains subnetwork 6 only. This indicates that the theme addressed in this subnetwork, is growing almost independently without much influence from other co-existing paradigms and it do not exercise influence over any emerging paradigms. The theme addressed is ‘nano positioning for precision mechatronic systems’. In the immediate future, this paradigm shift might influence many other applications that require precise positioning (e.g. manufacturing industry).

Fig. 15

Extended subnetwork showing subnetworks 7 and 17

Cluster #5

Cluster #5 consists of subnetworks 7 and 17, shown in Fig. 15. As found in FV gradient analysis, subnetwork 7 deals with the shift towards a paradigm in electronic circuit design for computing—‘nano electronic circuit design based on Quantum Dot Cellular Automata (QDCA)’. Subnetwork 8 seems to be a natural extension of this theme, and deals with the ‘Multi Stage Interconnection Network (MSIN) using QDCA’. As of now, subnetwork 7 is found to complement the developments in subnetwork 17, however, in future, innovative designs in 17 may drive technological development in 7.

Fig. 16

Extended subnetwork showing subnetworks 8 and 12

Cluster #6

Subnetworks 8 and 12 are found in this cluster, as shown in Fig. 16. Subnetwork 8 is found to deal with the ‘life cycle impact assessment of nano products’, especially predicting the impact of nano products based on the nano components. Subnetwork 12 deals with ‘nano ethics—the need for integration of Ethical Legal and Social Issues (ELSI) into science and engineering research related to nanotechnology’. As depicted in Fig. 16, these subnetworks show weak interaction and almost independent growth. So now, they can be treated as co-existing paradigms. But owing to the themes addressed, they have the potential to influence future developments of all other subnetworks.

Fig. 17

Extended subnetwork showing subnetwork 9

Cluster #7

In this cluster (shown in Fig. 17), there is only one subnetwork, subnetwork 9. As per the content analysis of this subnetwork, discussed during FV gradient analysis, this subnetwork addresses the theme ‘design optimization for nanotechnology based integration’ or simply ‘Nano Integration’. This can be viewed as a part of a major shift towards ‘nano computing’ paradigm.

Since there are no subnetworks which is directly connected to this subnetwork during its extension to immediate neighbourhood, its progress can be viewed as an independent one and therefore, the ‘nano integration’ paradigm is a co-existing one to all the other likely emerging paradigms. However, in future, it could drive developments in ‘nano electronics’ and also could influence ‘nano communication’.

Fig. 18

Extended subnetwork showing subnetwork 13

Cluster #8

This cluster consists of subnetwork 13 only. It is shown in Fig. 18. It deals with Environmental Impact Regulation for safe implementation of Nano technology, or in short ‘Nano Environmental Impact Regulation’, as used in Fig. 18.

This theme is visible as an independently growing team and is not competing or complementing any other paradigm. Therefore, currently this is a co-existing paradigm, but like ‘nano ethics’ and ‘Life cycle Impact assessment of Nano products’, it may have an influence on almost all other paradigms.

Fig. 19

Extended subnetwork showing subnetwork 14

Cluster #9

This cluster is shown in Fig. 19. It consists of subnetwork 14, which deals with the new likely paradigm ‘Nano technology based Implantation’ especially in orthopedics.

This is one of the biomedical paradigm shift due to the invasion of nanotechnology, however much different one from ‘nanotechnology based targeted drug delivery and/or imaging’. Thus, this is an independently developing paradigm (co-existing) and might remain so for the coming years too until it influence or get influenced by other biomedical paradigm shifts related to ‘nanotehnology’.

Fig. 20

Extended subnetwork showing subnetwork 16

Cluster #10

This cluster consists of subnetwork 14, as shown in Fig. 20. It deals with the theme ‘knowledge asset and intellectual property management for firms during radical change/transition towards nanotechnology’ or simply ‘Transition Model for Knowledge and Intellectual Property Management’.

As observed from Fig. 20, this is also an independently emerging paradigm and thus can be treated as a co-existing paradigm now. However, taking into the knowledge/technology management flavour of this paradigm, it may influence and regulate other existing as well as future paradigms.

Findings and policy implications

For academic as well as industrial researchers, identification of these paradigms that are existing and are likely to emerge is very important. This provide a direction to their research and even help to identify a research problem easily. National research institutes can also make use of these for research planning. Academic research funding agencies and in case of industrial research, investors and financial officers could assess the contemporary and futuristic relevance of proposals. Major findings from FV gradient analysis and extended network analysis are summarized below:

• 1. Major paradigm shits due to invasion of nanotechnology are observed in ‘electronics (for communication)’ and ‘biomedical applications such as targeted drug delivery’, steadily establishing the paradigms ‘nano electronics’ and ’nanoparticles based targeted drug delivery’ respectively.

• 2. In Nano electronic paradigms, Carbon Nano Tube (CNT) based nano electronics is challenged by Graphene based nano electronics. Most of the other paradigms are found to be mutual complementary except ‘Nano communication networks’.

• 3. ‘Nano communication networks’ that will lead to the realization of Internet of Nano things (IoNT) is currently visible as independently growing, but has the potential to drive developments in many of its co-existing paradigms and anticipated paradigms.

• 4. ‘Nano electromagnetics’ and ‘nano dielectrics’ etc. are also identified as paradigms in nanotechnology for engineering. While ‘nano dielectrics’ is found to be emerging as an independent paradigm ‘nano electromagnetics’ seems to complement ‘CNT based nano electronics’ and ‘Graphene based nano electronics’.

• 5. ‘QDCA based electronic circuit design’ and ‘Multi Stage Interconnection Networks based on QDCA’ are the paradigms (former complements latter) related to ‘nano electronic circuit design’. This may influence ‘nano computing’ paradigm in future.

• 6. ‘Nano integration for nano computing’ is a co-existing, independently developing paradigm that would be influenced in future by ‘nano electronic circuit design’ and may influence industrial activities in future.

• 7. In ‘nano particles based targeted trug delivery’, Tf-conjugation, Folic acid conjugation, Cholic acid conjugation etc. are the competing NP conjugation schemes of the biopolymers. Other strategies such as post conjugation strategy (for folic acid conjugation), Magnetic NP system and QD loaded NP system etc. are also found to co-exist or complement the conjugation paradigms.

• 8. Due to the autophagy of NPs that may inhibit the effect of nanomedicine, as identified in the early detected paradigm, an anomaly that challenges all the NP based drug delivery paradigms has surfaced. As a result, many emerging paradigms are envisaged in this direction.

• 9. ‘Nano technology based implantation’ is another co-existing paradigm that is independently evolving. It is part of a major shift towards ‘nano medicine’.

• 10. ‘Precision mechatronics based nano positioning systems’ seems to an independently developing co-existing paradigm that will be an asset to many industries including manufacturing industry.

• 11. Paradigms such as ‘nano ethics’, ‘nano environmental impact regulation’, ‘nano product life cycle impact assessment’, ‘Transition model for Knowledge and IP Management’, etc., are the co-existing paradigms that may govern the industrial practices when ‘nano technology’ take over the techno-economic progress.

Policy implications

This study reveals so many insights that could benefit a multitude of target groups that include researchers from academia and industry and various policy makers. Implications for policy making purposes is only mentioned in this article. Major policy implications that can be derived from this investigation and its findings are summarized in Table 3.

Table 3 Policy Implications for various target groups

Conclusion

Fascinated by its very potential to replace and improve many of the life and industrial practices, we have analysed the literature of ‘Nanotechnology’ for the research area ‘engineering’ during the last 25 years, from 1st January 1989 to 22nd December 2014 after creating a scientific citation network. Due to its potential to invade into the general purpose techno-scientific areas to earn a stature of ubiquity, we were interested to know about the co-existing paradigms in this field and their competing or complementary or co-existence with independent evolutionary nature. FV gradient, a recently proposed metric that utilize the ‘flow vergence effect’ to identify the pivot papers of a paradigm shift was employed for the same. When the papers that form FV gradient pairs were extracted, 18 subnetworks that consisted of 25 pairs were found. As we proceeded with the FV gradient analysis (content analysis of the pairs) of these subnetworks, along with their specific contribution, the themes represented by these were identified. A general idea about the nature of the ‘to be established’ paradigms were obtained. To substantiate these findings, interactions among these subnetworks were also investigated. For that these 18 subnetworks were extended to their immediate neighbourhood. A decision framework that uses two colours to distinguish pivot papers from its FV pair has been introduced to lay thumb rules for the identification of nature of paradigms based on the interactions of the subnetworks. As a result, it was found that most of the paradigms related to ‘targeted drug delivery’ are competing with each other. ‘Nano electronics’ also had its own share of competing paradigms in the form of ‘CNT based nano electronics’ and ‘Graphene based nano electronics’. ‘Nano electromagnetics’ seems to complement these competing ‘nano electronic’ paradigms and in future it may compete with ‘CNT based nano electronics’. A notable co-existing, independently evolving paradigm that would revolutionize human way of life is ‘Nanonetworks’ that may lead to the realisation of Internet of Nano things (IoNT). Policy implications of these and other major findings of this study for potential beneficiaries are also discussed. In methodological perspective, the early detection capability of the FV gradient is revealed in this investigation. As early detection is as good as prediction, the detection of paradigm shift that dealt with the autophagy induction of NPs can be regarded to have predictive implications. This served as a base for further exploration of the predictive power of the FV gradient method as done in Prabhakaran et al. (2018). Now other possible future works that can be pursued is discussed.

The citation network methodology developed in this work which includes the dichromatic framework can be readily applied on patent citation networks. Such an endeavour may serve as an additional validation of the methodology and could extricate more policy implications. A network (heterogenous) of scientific and patent literature might also be used to investigate the causal linkages between science and technology. Explorations of the subfields that encloses these paradigms like ‘nanonetworks’, ‘NP based targeted drug delivery’, etc. through citation network analysis can be considered for the extension of current investigation. Another potential extension would be methodological. As FV gradient weighted arcs form a signed weighted network, immense research opportunities are there. Some of these include cluster analysis after community formation, bibliographic coupling and co-citation network analysis of FV gradient weighted citation networks.