Disruptive Innovation in Geoenvironmental Sensing: Bringing Raman Spectroscopy to the Field Joe Sinfield Purdue University April 18, 2008 1
Outline: A model to describe innovation Applicability of this model to research Example of disruptive innovation in geoenvironmental sensing 2
Innovation is often considered in the context of technological breakthrough 3
We use many terms to characterize innovation, but few convey its true essence Discontinuous or radical vs. Incremental 1 Competence enhancing vs. Competence destroying 2 Architectural vs. Generational 3 Core vs. Peripheral 4 Modular vs. Interdependent 5 + + + 1 Dewar and Dutton, 1986; Ettlie et. al., 1984; Damanpour, 1996; 2 Tushman and Anderson, 1986; Anderson and Tushman, 1990; 3 Henderson and Clark, 1990; 4 Clark, 1985; Tushman and Murmann, 1998 ; 5 Baldwin and Clark, 2000; Schilling 2000 4
One model synthesizes these perspectives 5 Pace of Technological Progress Incumbents nearly always win Sustaining innovations Performance Performance that customers can utilize or absorb Disruptive Innovations Entrants nearly always win Time Source: Christensen and Raynor; The Innovator s Solution
Understanding this perspective highlights different paths to research success Cash flows Technological sophistication + - Good enough Offering 1 Program failure Next gen Offering 2 Good enough Offering 2 Disruptive approach Goal (traditionally defined by technical metrics) Spin-off Spin-off Traditional approach Time 6
These principles have been demonstrated many times in industry: Adoption of the Transistor Major Established Electronics Markets: Performance In academia we commonly drive toward radical sustaining innovation: Special Notice SN08-19: Materials with Novel Transport Properties established (MANTRA); Proposer's Day Workshop, DATE: March 20, 2008; vacuum tube Primary technical areas of interest include particle rejection manufacturers and chemical separation. Only novel technologies that improve water permeability one hundred-fold over existing systems will be investigated. Time Disruptive innovation: Transistors vs. vacuum tubes Path taken by Source:Innosight, LLC; http://news.sel.sony.com; www.kirainet.com/english/history-of-sony/; http://www.dself.dsl.pipex.com/museum/comms/mechamp/mechamp.htm 7
What if we instead asked, For what is our current capability good enough? What if fuel cells were used in power tools instead of automobiles? 1960 1995 2000 1998 What if photovoltaics were used to periodically power small appliances in developing countries instead of western homes? What if electric arc furnaces were used to melt down recycled ferrous scrap for re-bar instead of trying to make structural steel? Source: http://americanhistory.si.edu/fuelcells/pem/pemmain.htm 8
We can apply these same concepts to geoenvironmental sensing Fieldability Sensitivity Breadth CONCEPTUAL Well suited Cost Time Tolerable Poorly suited Reusability Disturbance Preparation Resolution In-laboratory analyses Probes/portable units Point sensors Reach We would like a system that better manages the tradeoffs inherent to geoenvironmental sensing, particularly to target in-situ analysis 9
Conceptually, Raman spectroscopy offers the potential to better address the in-situ challenge Raman is an optical spectroscopic technique Incident photon Scattered photon Before collision After collision Physical mechanism entails observation of photonmolecule collisions Time scale of 10-12 seconds Elastic collision - Rayleigh scattering Inelastic collision - Stokes shift S 0 Raman shift provides indication of vibrational energy Inelastic collision - Anti-Stokes shift Raman cross-section is proportional to 1/λ 4 S 0 S 0 The fundamental principles underlying the Raman technique point toward a capability that could be non-destructive, highly specific, very fast, and rich in breadth. 10
However, viewed through a sustaining lens, the technique is perceived to have weaknesses Raman weaknesses The equipment used to perform Raman analysis is typically very expensive The Raman phenomenon is inherently weak The usefulness of Raman is hindered by fluorescence Sustaining innovation rules Better = More power, resolution, and speed It is better to avoid fluorescence than to manage it Implication: In fluorescence inhibited settings, Raman is the tool of the few vs. the many 11
Fluorescence limited Raman spectroscopy sustaining vs. disruptive perspective Performance Sustaining trajectory: Avoid fluorescence or suppress its effects with costly cumbersome laboratory equipment $200-300K + FT-IR spectroscopy Dual wavelength excitation Ultrafast excitation Disruptive opportunity: Enable low cost and fieldability in natural settings, by giving up speed and resolution Time 12
We can pursue this opportunity by taking advantage of recent technological advances Compact spectrometers Miniature diode lasers Fieldable configurations Useful λ s Low cost Pulsed Compact Small form factor No cooling Low cost Telecom fiber and switching Fieldable configurations Low cost Compact Low cost, high speed DAQ Fieldable configurations Low cost Compact 13
Employing some of these technologies, we can change the rules Fundamental Raman system components Laser source Sample Monochromator Detector ACQ Attributes Sustaining High average power CW N/M IR Nonfluorescing media High resolution FFT Rapid full spectrum detection (CCD) LN cooled Continuously integrated voltage FFT signal deconvolution Gives Size Cost Media Gets Resolution Speed Sensitivity Disruptive Low average power Pulsed VIS Fluorescing samples Low resolution Scanning spectral coverage Slow monochromatic detection (PMT) TEC cooled Timeresolved photon counting Gives Resolution Speed Gets Size/portability Media Affordability 14
On this basis, we can envision a highly flexible insitu monitoring system Sensor docking unit Optical switch Optical fiber Monitoring points 15
Ultimately, this type of system would satisfy many of the objectives of in-situ analysis Fieldability Sensitivity Breadth CONCEPTUAL Well suited Cost Time Tolerable Poorly suited Reusability Disturbance Preparation Resolution Raman Spectroscopy In-laboratory analyses Probes/portable units Point sensors Reach 16
Following the disruptive pattern, we started with a good enough system Prototype time-resolved visible Raman spectroscopy system (1) 532 nm Q-switched microchip laser (2) Trigger photodiode (3) Beam splitter (4) Co-linear focus/filter probe (5) Sample chamber (6) Monochromator (7) Photomultiplier tube (8) 100-ps Total Time costdigitizer <$35,000 17
The system relies upon time-resolved photon counting to mitigate fluorescence effects Raman Fluorescence RhB doped benzene Intensity (counts) Volts 15 10 5 0 Integrated signal (1ms) SNR = N/A Laser PMT On Off Time (ns) Threshold Arbitrary units 100 Time-resolved photon counting (2ns) 996 cm -1 75 50 SNR ~ 15 25 0 950 1000 1050 Raman shift (cm -1 ) Total counts 18
The system balances tradeoffs GIVES Use of scanning monochromator slows spectrum acquisition Low cost / low average power laser requires time to build SNR Low cost spectrometer limits Raman line resolution GETS Applicability to multiple compounds Potential to assess multicompound mixtures Efficient signal averaging to enable low concentration tests Ability to suppress fluorescence Balance of sensitivity and resolution with VIS excitation Fieldable versatility with fiber Key question: For what application is the prototype technology likely to be good enough? 19
We looked for a market where the system s capability would be good enough In-lab environmental sensor Goal Fieldable timeresolved Raman system Technological sophistication Edible oil analyzer Nitrate threshold detector Time 20
With improvements implemented through work on oils, system applicability has expanded Arbitrary units 1200 1000 800 600 400 200 0 Neat TCE Neat benzene 996 cm -1 N-O line 1041 cm -1 TCE 1590 cm -1 5 ppt aqueous NH 4 NO 3 solution 3063 cm -1 Water band 0 500 1000 1500 2000 2500 3000 3500 4000 Raman shift (cm -1 ) NH 4 band 21
The system has evolved quickly and is now amenable to a range of possible applications Low-cost, Time-resolved Raman System Geoenvironmental sensing Precision agriculture Food science Biochemistry 22
Concluding thoughts There are benefits to addressing the needs of the overshot Low end disruption does not necessarily mean low tech Democratization can drive improvement and new research direction Technology (and science) can improve faster through trial Advances occur when you engage more minds outside your field Don t let perfection be the enemy of the good enough -Meg Whitman Former CEO, ebay 23